Post on 24-Aug-2020
Supervisor: Rick Middel Master Degree Project No. 2015:51 Graduate School
Master Degree Project in Logistics and Transport Management
Forecasting Process for Predicting Container Volumes in the Shipping Industry
Solmaz Darabi and Mirza Suljevic
Abstract
Abstract In an industry that is fast moving, a company’s ability to align to market changes has
becoming a major competitive factor. Forecasting of future outcomes has thus become a
necessity for companies to prepare for uncertainties. The shipping industry is an industry
characterized by financial turbulence and ever- shifting demand. In order to adapt to
changing trends and enhance operational management it is therefore essential for companies
to implement proper forecasting processes. By understanding and implementing a well
functioning forecasting process companies can increase their forecast accuracy, thus reduce
their stock outs and increase their customer satisfaction.
The purpose of this paper was to evaluate the existing forecasting process at company X in
order to identify and propose an improved forecasting process for predicting container
volumes. The research was based on a case study, where the aim was to create a detailed and
in- depth understanding of the subject. To identify the answer for the research question
various forecasting processes suggested by the literature have be investigated. Based on
presented literature a new forecasting process has been created and suggested for
implementation by company X. The implementation of a forecasting process is essential for
company X in order to adapt to continuously changing trends, improve their performance
and strengthen their competitive position. The design of a new forecasting process for
predicting container volumes will allow company X to reach sustainable results.
Acknowledgement
Acknowledgement With this acknowledgement, we would like to express our gratitude to company X for
enabling us the opportunity to do our Master thesis for the company. We would like to thank
all employees at company X for their contributed time and effort in supporting us and
providing valuable information.
We would also like to sincerely acknowledge our supervisor Doctor Rick Middel at the
school of Business, Economic and Law at the University of Gothenburg, for his guidance,
supervision and support throughout the entire process.
Finally, we dedicate this work and give a special thanks to our loved ones - family and
friends for all the encouragement they have given us during the entire process of this thesis.
List of figures
List of Figures Figure 1. Disposition of the thesis
Figure 2. Shima and Siegel`s forecasting model
Figure 3. Brockwell and Davis`s forecasting process
Figure 4. Schultz`s forecasting model
Figure 5. Winklhofer, Diamantopoulos and Witt`s model
Figure 6. A proposed forecasting process based on the existing literature
Figure 7. Conceptual model of judgment
Figure 8. The forecasting system
Figure 9. A summary of the literature review
Figure 10. Information distribution design at company X
List of Tables Table 1. A comparison of the forecasting processes
Table 2. Forecast of 40 HC demand using moving average & exponential smoothing method
Table 3. Evaluation of the forecasting results
Table 4. A comparison between company X `s and the proposed forecasting process
List of Graphs Graph 1. Available container volumes, Sweden 2013 & 2014.
Graph 2. Export bookings, Sweden 2013 & 2014
Graph 3. Import volumes, Sweden 2013 & 2014
Graph 4. Simple linear regression technique for 40 HC, Sweden 2014
Abbreviations
Abbreviations Abbreviation Full name
DC Dry cargo
HC High cube
KPI Key performance indicator
MAD Mean absolute deviation
MAPE Mean absolute percentage error
MSE Mean square error
SES Simple exponential smoothing model
TEU Twenty foot equivalent unit
Table of content
6
Table of Contents
ABSTRACT 2
ACKNOWLEDGEMENT 3
LIST OF FIGURES 4
LIST OF TABLES 4
LIST OF GRAPHS 4
ABBREVIATIONS 5
1. INTRODUCTION 8 1.1 BACKGROUND 8 1.2 PROBLEM FORMULATION 10 1.3 PURPOSE 11 1.4 RESEARCH QUESTION 11 1.5 LIMITATION 11 1.6 DISPOSITION OF THE THESIS 12
2. LITERATURE REVIEW 14 2.1 THE CONCEPT OF FORECASTING PROCESS 14 2.2 SHIMA AND SIEGEL`S FORECASTING MODEL 15 2.3 BROCKWELL AND DAVIS`S FORECASTING PROCESS 15 2.4 SCHULTZ`S FORECASTING MODEL 16 2.5 WINKLHOFER, DIAMANTOPOULOS AND WITT`S MODEL 17 2.6 IDENTIFY STRATEGIC GOALS 21 2.7 DECIDE THE OBJECTIVES OF THE FORECAST 22 2.8 CHOOSE A FORECAST APPROACH 22 2.9 CHOOSE FORECAST VARIABLES FROM COLLECTED DATA 23 2.10 CHOOSE FORECAST HORIZON 24 2.11 IDENTIFY DEMAND PATTERN 25 2.11.1 DEMAND PATTERNS 25 2.11.2 DECOMPOSITION OF DATA 28 2.12 CHOOSE FORECASTING TECHNIQUE 31 2.12.1 QUANTITATIVE METHODS 32 2.12.2 CASUAL METHODS 37 2.12.3 BENEFITS AND DRAWBACKS OF DIFFERENT FORECASTING TECHNIQUES 39 2.13 EVALUATION OF FORECASTING TECHNIQUE AND RESULT 40 2.13.1 FORECAST ACCURACY, ALSO KNOWN AS ERROR MEASUREMENT 40 2.13.2 EVALUATION OF FORECASTING RESULTS 42 2.14 INTEGRATION OF FORECASTING PROCESS 43 2.15 SUMMARY OF THE LITERATURE REVIEW 46
3. METHODOLOGY 47 3.1 RESEARCH APPROACH 47 3.2 RESEARCH DESIGN 47 3.3 RESEARCH STRATEGY- MIXED METHODS 48
Table of content
7
3.4 RESEARCH ONTOLOGY- INTERPRETIVISM AND POSITIVISM 49 3.5 CLASSIFICATION OF THE RESEARCH BASED ON LOGIC OF THE RESEARCH 51 3.6 METHOD OF DATA COLLECTION 51 3.6.1 PRIMARY DATA 51 3.6.2 INTERVIEW PARTICIPANTS 52 3.6.3 INTERVIEW METHOD AND STRUCTURE 52 3.6.4 SECONDARY DATA 53 3.7 RESEARCH EVALUATION- RELIABILITY AND VALIDITY 53
4. EMPIRICAL FINDINGS 56 4.1 THE CASE COMPANY 56 4.2 INFORMATION DISTRIBUTION DESIGN 56 4.3 COMPANY X`S FORECASTING PROCESS 57 4.3.1 VARIABLE: AVAILABLE CONTAINERS 58 4.3.2 VARIABLE: BOOKINGS 60 4.3.3 VARIABLE: IMPORT VOLUMES 62
5. ANALYSIS 64 5.1 THE IMPORTANCE OF A FORECASTING PROCESS 64 5.2 THE FORECASTING PROCESS OF COMPANY X 64 5.2.1 IDENTIFY STRATEGIC GOALS 64 5.2.2 DECIDE THE OBJECTIVES OF THE FORECAST 65 5.2.3 CHOOSE A FORECAST APPROACH 66 5.2.4 CHOOSE FORECAST VARIABLE FROM COLLECTED DATA 66 5.2.5 CHOOSE FORECAST HORIZON 67 5.2.6 IDENTIFY DEMAND PATTERNS AND DECOMPOSITION OF DATA 67 5.2.7 CHOOSE FORECASTING TECHNIQUE AND ERROR MEASUREMENT 68 5.2.8 EVALUATION OF FORECASTING TECHNIQUES & RESULTS 71 5.2.9 INTEGRATION OF FORECAST PROCESS 72 5.3 COMPARISON BETWEEN COMPANY X `S AND THE RESEARCHERS PROPOSED FORECASTING PROCESS 73
6. CONCLUSION 74 6.1 FUTURE RESEARCH SUGGESTIONS 77
7. REFERENCES 79
APPENDIX 85
Introduction
1. Introduction In this chapter the authors will present the background of the thesis subject. Continuing, a problem formulation is defined followed by the purpose of the study. Thence, a main research question is formulated that gives further direction to the author’s interest. The introduction will be concluded with the limitation of the thesis in order to define the scope of the study.
1.1 Background The use of containers in sea and ocean transportation has gradually increased since their
introduction half a century ago (Beenstock and Vergottis, 1993). This growth has been
enhanced even more by economic development and globalization trends during the last
couple of decades (McKinnon, 2010). Global trade and financial turbulence have created an
industry that is highly volatile, uncertain and if not predicted accurately, can cause financial
instability for organizations. In order to adapt to ever changing trends and enhance
operational management it is therefore essential for companies to implement proper
forecasting processes. By understanding and implementing a well functioning forecasting
process companies can increase their forecast accuracy, thus reduce their stock outs and
increase their customer satisfaction (Jacobs, Chase and Lummus, 2011).
A good forecasting process is central for daily operational management and vital for every
significant management decision, as it eases business planning and makes it more efficient
(Diaz, Talley and Tulpule, 2011). The objective is to provide a continuous flow of
information, hence enabling organization to cope with the ever-changing shift in demand
and supply, increase operational efficiency and manage and mitigate risk within a market
(Vlahogianni, Golias and Karlaftis, 2004). The aim of a forecasting process is to provide its
executives and management with a proper tool to improve their performance and
competitive position while adjusting to rapid changes in the economy (Pal Singh Toor and
Dhir, 2011). By designing a forecasting process that aligns with strategic goals, a company
can use the forecasting process as a mean of sustaining competitive advantages (Daim and
Hernandez, 2008).
By embedding the forecasting process in the organizational decision making process, a
clearer picture of the forecasting contribution to organizational effectiveness will emerge.
Introduction
9
The construction of a process that reflect a realistic assessment of current business
environment can help companies prepare and respond to dynamic market situation, thus
increase effectiveness and competitive advantage (Gardner, Rachlin and Sweeny, 1986).
Conducting a forecasting process that is based on the most important key performance
indicators for effective performance of a company will support improvements of the
business process (Janes and Faganel, 2013). Hence, forecasting processes should align with
the strategic goals of an organization. Companies that don’t align their forecasting process
with strategic planning of their operations can experience lack of clarity regarding the
structure and responsibility, and miss opportunities to reallocate resources and take
advantage of market opportunities (Makridakis and Wheelwright, 1982).
Today forecasting consists of several complex disciplines. Some methods aim at identifying
the underlying reasons that might influence the variable that is being forecast, while other
techniques incorporate judgments and opinions. Each method has a special use, and much
consideration must be placed in selecting the most appropriate technique for a particular
application. The aim is to identify a forecasting process that will generate highest level of
accuracy. Still, there are some who question the reliability and validity of the forecasting
discipline (Jarrett, 1987). Various organizations hold a false belief that the future holds
enough time to allow organizations to react to a change in events, or, they believe that the
future holds no important change (Jarrett, 1987). According to Diaz, Talley and Tulpule
(2011), improved forecasting processes will have a direct and positive impact on several
aspects of an organization. The integration of forecasting processes across an organization
will enhance coordination in establishing plans consistent with corporate strategy, hence
improve organizational alignment and financial performance (Pal Singh Toor and Dhir,
2011). A good forecasting process can fail through poor integration. A well- integrated
forecasting process is therefore a necessity in order to enhance the results of the process.
While many researchers have explained the disciplines of forecasting methods (Daim and
Hernandez, 2008), the incorporation of forecasting process in managerial decisions has
received little attention. The aim of this thesis is to explain how forecasting as a sustainable
process should be conducted and incorporated as a decision support tool, in order to increase
accuracy in future decisions. A case study based on company X is presented.
Introduction
10
1.2 Problem formulation According to Stopford (2009) “the problems of making decisions about an uncertain future
are as old as the shipping industry”. The maritime industry has long struggled with making
accurate forecasts, partly due to different aspects of the industry that are problematic to
predict. One example to illustrate difficulties with accuracy is through predictions of future
freight rates. Freight rates are much dependent on the quantity of ships being ordered, a
behavioural variable which is affected by shipping cycles and development in world
economy. These variables are extremely complex to predict, hence making it difficult to
forecast accurately. The ability to anticipate market movements has long been deficient,
despite attempts to develop efficient forecasting techniques. Still, this does not imply that all
forecasting attempts are set to fail, rather it’s an indication that the shipping industry is a
complex industry to predict (Stopford, 2009).
While many researchers have explained the disciplines of forecasting methods (Daim and
Hernandez, 2008), the subject of forecasting processes has received little attention. Existing
literature does not acknowledge any forecasting process that takes into consideration all
necessary steps that are required to conduct accurate forecasts. A sustainable forecasting
process is an on-going process that enables forecasters to conduct improved predictions by
evaluating, monitoring and refining forecasts through time. Forecasting, as an on-going
process should provide a basis for future predictions, allowing a forecaster to make
appropriate adjustments that align with the organisations objectives and reality, thus making
continuous improvements (Ord and Fildes, 2013). The authors of this study have
investigated an extensive range of literature and identified a gap regarding forecasting
processes that needs to be satisfied. Existing literature confer how accurate forecasts can be
generated by using different forecasting techniques. However, little attention has been
placed on forecasting as a sustainable process. The authors of this thesis believe that there
are several aspects of forecasting in the shipping industry that goes beyond merely the use of
forecasting techniques. The aim is thereof to recapitulate these aspects in one single
forecasting model in order to increase accuracy of future predictions.
At current state company X is struggling with their forecasting activities of available
containers. The aim of the company is to always satisfy all container bookings, while
staying within the margins provided by the head office. Yet, due to ineffective forecasting
activities this has become an extensive struggle. The inability to make accurate forecast of
Introduction
11
customer demand has caused loss of bookings as well as a surplus of longstanding
containers. Ineffective forecasting have lead to difficulties meeting customer demand, long
waiting time for receiving containers and decrease in service level. This has accordingly
created a request from company X to identify a forecasting process that will enable them to
satisfy all bookings, minimize costs and increase service-level.
1.3 Purpose The purpose of this paper is to evaluate existing forecasting activities implemented by
company X, in order to identify and propose an improved forecasting process for predicting
container volumes. Thereby, allowing company X to experience increase forecast accuracy.
Various forecasting processes suggested by theories will be investigated in order to identify
similarities, strengths and shortcomings that each chosen forecasting process posses. The
existing theories will thus enable the researchers to construct a forecasting process, which
will be used as a basis for further research of the topic. Each step of the constructed
forecasting process will be evaluated. Several different forecasting techniques will be
included and compared in order to identify different contributions and shortcomings. In
addition a comparison and evaluation of the magnitude of forecasting error will be
conducted. By analysing different forecasting processes, the authors expect to have the
means to suggest an improved forecasting process that can be incorporated as a decision
support tool in order to increase accuracy in future decision. Thereby achieve the overall
purpose of the paper.
1.4 Research question The following research question will be investigated in order to propose a possible solution
to the problem:
“How should a forecasting process for predicting container volumes at company X be
designed, in order to generate accurate forecasts?”
1.5 Limitation A case study of company X has been conducted, with the geographical focus on company X
in Sweden. In order to narrow down the scope of the research, the analysis was based only
on quantitative and casual methods, excluding qualitative forecasting methods from the
research scope. Qualitative models are based on judgmental forecasting and are
implemented in situations where pure statistical methods are not possible due to lack of
historical or economical data. The technique is beneficial when other methods are not
Introduction
12
adequate (Rowe and Wright, 1999). Since historical data provided from company X were
merely quantitative, qualitative models was not considered sufficient due to their subjective
characteristics. The authors of this research have focused on explaining moving average
methods, exponential smoothing methods and regression analysis. However, only the
techniques of simple moving average, Holt-Winters non-seasonal exponential model and
simple regression analysis were tested. Regarding decomposition of time series, only
software R was used, as it was believed to be the most efficient way to conduct the research
and extract the main components of the empirical data. Furthermore, not all- statistical error
measurement was tested. The authors have focused on MAD, MSE and MAPE.
Moreover, the aim of the research was merely to present a forecasting process that is
believed to be effective for future forecasting practices. No efforts were put on
implementation of the forecasting process in company X. A further limitation was that only
20-40 DC / HC has been subject of analysis. It is also important to mention that no
measurements regarding how an improved forecasting process would affect profits, revenues
and costs have been conducted.
1.6 Disposition of the thesis
Figure 1. Disposition of the thesis
CHAPTER 1 Introduction
CHAPTER 2 Literature review
CHAPTER 3 Methodology
CHAPTER 4 Emperical findings
CHAPTER 5 Analysis
CHAPTER 6 Conclusion
Introduction
13
Chapter 1. Introduction
In chapter 1 the authors have presented a general introduction to the background of the
research subject, followed by a description of the problem background and a problem
formulation. Furthermore, a purpose statement, research question and an explanation of the
thesis limitations are outlined in order to create an understating of possible restrictions.
Chapter 2. Literature review
The literature review is an assessment of existing scholarly material addressing the research
question of this thesis. The literature review starts by presenting a number of forecasting
processes followed by a comparison. Subsequently, each step within the process is presented
and clarified.
Chapter 3. Methodology
In this chapter the methods derived to solve the research question has been defined. The
research approach, research design, research strategy, methods of data collection and
research evaluation has been clarified and outlined.
Chapter 4. Empirical findings
The objective of this chapter is to present the case study. More specifically how the
company conducts their forecast process in relation to their operations.
Chapter 5. Analysis
In chapter 5 an analysis has been conducted by comparing the literature review to the
empirical findings. The analysis can thereafter be used to form the basis for a conclusion.
Chapter 6. Conclusion
The results of the thesis are developed and presented in this chapter. Furthermore, the
authors provide suggestion for improvement for the case company, as well as a proposal for
future research.
Literature review
2. Literature review In this chapter the authors will present literature relevant to the subject of forecasting process. A comprehensive frame of reference is presented that that will increase the readers understanding of the subject. Numerous forecasting processes are explained and compared followed by a detailed description of each step of a forecasting process.
Literature regarding forecasting is vast and diversified, where the biggest attention has been
paid to different forecasting techniques and how they can be combined in order to improve
the accuracy of the forecasting process. Much research has not been done that goes beyond
developing and testing various forecasting techniques, shaping a big gap between
application of forecasting techniques and their development (Schultz, 1992). The literature
review starts by introducing the concept of forecasting and its importance. Continuously, an
explanation and comparison of existing forecasting processes have been conducted. In
addition, a new forecast process has been constructed, which is a summary of the literature
and the presented forecasting processes.
2.1 The concept of forecasting process Forecasting process is an essential activity in many business areas. It eases the determination
and adaptation to future demands, allowing a company to reach sustainable solutions and
growth opportunities. The desire to forecast rises from the need of predicting future
economic conditions and the wish to eliminate future uncertainties and risks (Thomopoulos,
1980). The forecasting discipline is defined as “a planning tool that helps management in its
attempts to cope with the uncertainties of the future, relying mainly on data from the past
and present and analysis of trends” (BusinessDictionary.com, 2015).
Forecasting as a process is a vital part of business organization as it provides the basis for
planning and decision- making (Jacobs, Chase and Lummus, 2011). The process has become
a necessity for management to cope with increasing complexity of managerial forecasting
problems and rapid changes in the economy (Thomopoulos, 1980). The implementation of a
well- constructed forecasting process can help companies improve their performance and
competitive position as well as plan tactics to match capacity with demand, thereby
achieving high yield levels. By implementing a tool that generates up to date information,
companies are able to take better advantage of future opportunities and reduce potential risks
Literature review
15
(Daim and Hernandez, 2008). The success of a forecast lays heavily on accuracy, if level of
accuracy is low a forecast can be extremely misleading, causing costly damages. It is
therefore of great importance to monitor forecast errors to validate that errors are within
reasonable boundaries. However, the complex nature of the world economy can make it
difficult to predict future values for various variables, in spite of sophisticated mathematical
models. A great deal of visibility, information sharing and communication between different
divisions of a company is therefore a necessity (Granger and Pesaran, 2000).
2.2 Shima and Siegel`s forecasting model Shim and Siegel (1999) present a simple forecasting process consisting of six basic steps.
The framework of this forecasting process starts with determination of what is to be
forecasted, indicating the required level of details. Before selecting a forecasting method it is
necessary to establish a time horizon. According to Shim and Siegel (1999), the process of
gathering the data and developing a forecast occurs after the selection of a forecasting
method, followed by identification of any assumptions that should be made in order to
prepare the forecast. Monitoring the forecast is named as the final step of this forecasting
process, where an evaluation system is considered as a matter of choice.
Figure 2. Shima and Siegel`s forecasting model (1999)
2.3 Brockwell and Davis`s forecasting process Brockwell and Davis (2010) present a general approach to time series modelling that
consists of several steps. In the first step of a forecast process, data features should be
examined by using various plotting techniques, particularly checking for trend, seasonal and
other components of the data. Furthermore, the analyst should remove trend and seasonal
Decide the purpose and what is to be forecasted
Establish a time horizon
Select a forecasted method
Gather appropriate data
Make forecasts
Monitor forecasts
Literature review
16
components from the data in order to get stationary residuals. Continuously, a forecasting
model should be chosen based on the residuals generated in the previous step. An important
part of Brockwell and Davis`s forecasting process is the prediction of right residuals and
inverting them into previous steps in order to arrive at forecasts of the original series.
Brockwell and Davis (2010) mention an alternative briefly explained approach, where time
series are expressed in terms of sinusoidal waves of different frequencies also known as
Fourier components.
Figure 3. Brockwell and Davis`s forecasting process (2010)
2.4 Schultz`s forecasting model Contradictory to Brockwell and Davis (2010), Schultz (1992) places a lot of emphasis on
how forecasting techniques can be implemented into fundamental processes of strategic
planning and assessment. Instead of focusing on forecast technique evaluation, model
building and testing, the fundamental aspects of forecasting in organizations should be in the
areas of decision making and policy making processes, where implementation and
evaluation of forecasts impacts on a corporate level. Thus, questions such as whether
strategic goals can affect forecasts arise. Continuously, the gap between the development of
various forecasting techniques and their implementation is in the centrum of discussion.
Differing from many other forecasting models, process of forecast evaluation is based on
objective measures such as sales, costs and profit, as at the end of the day this is the only
thing that counts.
Plot the data
Decomposition of the data in order to get stationary residuals
Choose a model to fit the residuals
Compute forecasting by predicting right residuals
Use alternative approach or express the data in form of Fourier components
Literature review
17
Figure 4. Schultz`s forecasting model (1992)
2.5 Winklhofer, Diamantopoulos and Witt`s model Similar to Schultz (1992), Winklhofer, Diamantopoulos and Witt (1996) developed a
framework of forecasting processes that helps forecasting experts and decision makers to
understand and to use different forecasting techniques. Their framework is used as a strong
bridge between theory and practice and it is an extension to the previously explained
Schultz`s (1992) model. The model explains three sets of issues linked to, design,
selection/specification and evaluation of forecasting. Design issues include important
enquiries regarding purpose and use of forecast, where forecast level, time horizon,
resources, and forecast users are the main areas of the analysis.
Furthermore, Winklhofer, Diamantopoulos and Witt (1996) explain selection/specification
concerns as set of issues related to awareness of various forecasting techniques and
identification of main factors that affect forecasting technique selection. The final part of
Winklhofer, Diamantopoulos and Witt`s (1996) framework for organizational forecasting
practice is evaluation activities, where great emphasis is put on the implementation and
presentation of forecasts to management. Moreover, the issues related to standards for
forecast evaluation, performance measurement and forecast improvement are raised as a
complex and biased parts of the evaluation section. Finally, concerns regarding integration
of involved parties in the forecasting process and information sharing between different
parts of the supply chain or organizational levels are described rather as a necessity than a
matter of choice.
Decision making / policy making process
Implementation (Model building and testing)
Forecast evaluation
Literature review
18
Figure 5. Winklhofer, Diamantopoulos and Witt`s model (1996)
DESIGN ISSUES Purpose / use of forecast Forecast level Time horizon and frequency of forecast preparation Resources commited to forecasting Forecasting preparers Forecast users Data sources
SELECTION /SPECIFICATION ISSUES Familiarity with forecasting techniques Criteria for technique selection Usage of alternative forecasting methods
EVALUATION ISSUES Forecast presentation to management Forecast review and use of subjective judgement Standards for forecast evaluation Forecast performance Forecasting problems and forecast improvement
Literature review
19
A comparison of explained forecasting processes are presented in the table below.
Model
Shim & Siegel
Brockwell &
Davis
Schultz
Winklhofer, Diamantopoulos
& Witt
Strategic goal alignment
✓
Decide the objective
✓
✓
Time horizon
✓
✓
Data collection process
✓
✓
Decomposition of data
✓
Select a forecast technique
✓
✓
✓
Make the forecast
✓
✓
✓
✓
Evaluation of results
✓
✓
✓
Monitoring the forecast
✓
✓
✓
Use alternative approach
✓
✓
Integration of results
✓
Table 1. A comparison of the forecasting processes In table 1 a summary of existing forecasting processes and the areas that each process
focuses on is presented. It can be seen from the table that no single model includes all
above- mentioned steps. This inspired the authors to compute a more comprehensive and
satisfactory forecasting process that may be incorporated as a decision support tool. The
table is a verification of the theoretical contribution of this thesis, as it clearly identifies the
research gap in existing literature. Emphasis is placed on each step of the process, as each
Literature review
20
step is considered highly important. As mentioned before, much of the existing literature
focuses on improving different forecasting techniques, while some basic steps from the
world of business development were ignored. The proposed forecasting process focuses
equally on every step, starting with identification of corporation`s strategic goals as it gives
clear indications on which fields a corporation should compete on (Makridakis and
Wheelwright, 1987). Moreover, every single step in this process is carefully analysed,
evaluated and placed so it eases the entire process and makes it sustainable, enabling a
corporation to experience continuous improvements.
From the table it can be read that Winklhofer, Diamantopoulos and Witt`s forecasting model
is the most comprehensive compared to the other three models, still the model is missing
some important steps such as strategic goal alignment and decomposition of the data. By
combining the most essential steps, the authors were able to cover all the significant parts of
a forecasting process. By integrating existing models into one and adding to it additional
knowledge from the world of forecasting, a new forecasting process was formed and used as
a basis for the further studies.
Figure 6. A proposed forecasting process based on the existing literature
IDENTIFY STRATEGIC GOALS
DECIDE OBJECTIVES OF THE FORECAST
CHOOSE A FORECAST APPROACH
CHOOSE FORECAST VARIABLES FROM COLLECTED DATA
CHOOSE FORECAST HORIZON
IDENTIFY DEMAND PATTERN
CHOOSE FORECAST TECHNIQUE
EVALUATION OF FORECASTING TECHNIQUE & RESULTS
INTEGRATION OF FORECAST PROCESS
Literature review
21
2.6 Identify strategic goals Importance of forecasting in planning and decision-making process has become apparent in
areas of business sustainability and development. Organizations have moved towards a more
systematic way of making decisions, where explicit justifications are necessary for their
implementations and forecasting is one way in which these activities can be supported
(Winklhofer, Diamantopoulos and Witt, 1996). Therefore, one of the primary issues in
managing development and use of forecasting processes is careful mutual consideration of
criteria to be satisfied (Makridakis and Wheelwright, 1987).
Good forecasting processes should be constructed in response to the policy and in alignment
to strategic goals of an organization, so that expected trade offs between customer service
level, inventory levels and procurement economics can easily be interpreted. These trade-
offs are fundamental to the organization's survival, as customer service level must be high
enough to satisfy customer`s requirements, while taking into consideration constraints of
inventory levels and cost efficiency (Makridakis and Wheelwright, 1987). Good analytical
input is required for effective decision making at policy level. Planners need to ensure that
top management has adequate understanding (Mintzberg, 1976). Therefore, embedding the
forecasting process in the organizational decision making process is crucial and can improve
organizational effectiveness as well as enhance its competitive advantage (Schultz, 1992).
Moreover, Gardner, Rachlin and Sweeny (1986) state that forecasting should be based on
the competitive ideas of how and where the business should compete, which means that
forecasting process should be developed after, not before, these competitive ideas are
outlined. Additionally, in order to be meaningful, forecasts should be developed with the
participation of the people who really understand the company`s competitive strategy, as the
real value of forecasts is in understanding future opportunities (Gardner, Rachlin and
Sweeny, 1986).
In practice organizations use key performance indicators (KPIs) in order to monitor the
fulfilment of the organization`s strategy and their competitive advantage (Janeš and Faganel,
2013). KPIs reflect organization`s goals, which means that if an organization has the goal of
being the most profitable organization, then it will have a KPI that measures profitability
(Shahin and Mahbod, 2007). It is therefore of great importance to relate the forecasting
process to the organization`s KPIs. Janes and Faganel (2013) state that diagnostic activities
Literature review
22
such as forecasting of future demand should be based on the most important KPIs in order to
support improvements of the business process. According to Shahin and Mahbod (2007),
goal setting is one of the first steps an organization needs to complete. Thus, aligning the
forecasting goal to the strategic goals of an organization and its KPIs is the first step that
forecast preparer should take when conducting a forecasting process.
2.7 Decide the objectives of the forecast One of the initial steps of every forecasting process is to define what is to be forecasted,
identify objectives and key elements that need to be considered and forecasted. By defining
the problem one will create an understanding for the intended use of the forecast and how
the forecasting purpose fits within the organization requiring the prognoses (Daim and
Hernandez, 2008). This will subsequently indicate the level of detail, units of analysis and
time horizon required in the forecast (Shim and Siegel, 1988). A forecast can be either
normative or informative in nature. A normative forecast is goal oriented and refers to the
needs of an organisation. Normative forecasting takes into account an organization's
purpose, mission and expected achievements (Daim and Hernandez, 2008), while an
informative forecast provides information regarding market disruption (Winklhofer et al.,
1996).
Deciding on the objectives of a forecast helps organizations choose a forecasting technique
with an appropriate level of sophistication. For example, if the intention of the management
team is to forecast effects of a specific strategy, the chosen technique must be sophisticated
enough to explicitly take into account all aspects that may affect the underlying strategy.
Level of accuracy is therefore of great importance and needs to be defined, or in other
words, the level of inaccuracy that is acceptable. The level of tolerable in accuracy allows
the company to trade off cost against the value of accuracy in choosing a technique. This can
be illustrated through an example. In production and inventory control increased accuracy
most often leads to lower safety stocks. The manager must therefore weigh the cost of a
more sophisticated technique against potential savings in inventory costs (Chambers,
Mullick and Smith, 1971). Once the purpose of the forecast is defined, the forecast approach
needs to be determined.
2.8 Choose a forecast approach There are two general approaches when doing forecasting:
Literature review
23
1. Exploring historical data
2. Exploring other factors that could affect forecasting process (Axsäter, 2006).
When analysing historical data, a forecasting process is conducted in regard to the previous
demand data or time series (Axsäter, 2006). Time series can be explained as a set of
measurements that are ordered over time, possessing special characteristics associated with
the sequence of the observation (Newbold, Carlson & Thorne, 2013). Forecasting techniques
used for this approach are based on statistical methods for analysis of time series. This
approach is easily applicable to operational control systems, as the forecast process can be
updated for huge number of items on regular basis (Axsäter, 2006).
In regard to the second approach, it is very usual that a forecast of demand for a certain item
is based on demand for another item or variable (Axsäter, 2006). Jacobs, Chase and
Lummus (2011) name this type of demand as the dependent demand. A good example is an
item that is used solely as a component for manufacturing a final product. In order to make
an accurate forecast for the item, it is necessary to forecast the demand for the final product
first. Demand for certain products can be related to different variables. For example,
forecasting of demand for ice cream sales is related to the weather forecast. One of the most
common variables that affect demand of a product is price, but it can also be others, such as
advertising expenditure from the previous months or gross domestic product (Axsäter,
2006). This kind of approach explains functional relationship between two or more
correlated variables, where one of them is used to predict the other one (Jacobs, Chase and
Lummus, 2011).
2.9 Choose forecast variables from collected data The quality of a forecast is only as good as the quality of the input data. It is therefore of
great importance to ensure that all bad and irrelevant data is filtered out in order to increase
the accuracy of the forecast. Low quality data may be a result of bad data collection process
or the use of outliers that don’t represent general trends. The availability of data in
conjunction with the correctness of the data affects the accuracy of the forecast. By
collecting related input data from different sources, the trustworthiness and reliability of the
information will increase (Malakooti, n.d.).
A proper selection of input variables is necessary to build an appropriate model with high
performance. The objectives are to find an optimal set of variables. Too many variables can
Literature review
24
cause the model to become over fitting, and cloud the true relationship between existing
variables, while too few variables might not be enough to capture the dynamics of the
phenomena of study. Choosing a set of relevant variables can be done through the use of
algorithms. However, most forecasting models consist of a specific formula where it is
clearly stated what variables to include in the model. If the correlation between variables is
high, the forecast will generate high accuracy. Conversely, if the correlations between
variables are low, one will generate a forecast with lower accuracy (Tran, Muttil and Perera,
2015).
2.10 Choose forecast horizon Time is a vital element of planning, decision- making and implementation as it improves the
competitive performance of a company. The time horizon of a forecast is the estimated
length of time that a company decides to predict. The time horizon of a forecast has
implications for the choice of forecasting method and model construction (Cook, 2006). The
forecasting time horizon can be divided into short- term, medium-term and long term. Since
forecasts are crucial components in a corporation’s decision- making process it must be
produced and received within a correct time frame. Since, a forecast is of little use if it’s
received too late. Moreover, the time frame must include information regarding how far in
the future the forecast will be done, how far in the past the data is relevant, in addition to,
how much time is available and required to employ the forecasting method. When deciding
on the time period that is to be covered, one needs to keep in mind that accuracy decreases
as the time horizon increases (Malakooti, n.d.).
Short- term forecast also known as tactical forecast, range up to one year. Short- term
forecast is employed to adjust an existing plan based on new information obtained. The data
generated from the forecast are used as input for tactical decisions. Errors that occur during
this time frame are more due to random events, and less a result of cyclical and seasonal
patterns. Short- range forecasts are typically the most accurate (Malakooti, n.d,).
Medium- term or intermediate range forecast includes a planning horizon between one and
three years. Medium- term forecasts are often used as a starting point to annual business
planning. Since the time horizon is considerably longer than short- term forecasts, cyclical
and seasonal patterns have a significant impact and must be considered in the analysis. The
information generated during this time period is used for tactical decisions, to somewhat
strategic decisions. Use of regression based methodology and extrapolative methods are
Literature review
25
often employed in medium- term forecasts (Malakooti, n.d.).
Long term forecasts or strategic forecasts, concern purely strategic decisions ranging from
three years and up. Strategic forecasts are exposed to threats of long- term trends and
business cycles. The level of uncertainty in long- term forecasts tends to be high. As the
time horizon increases the more inaccurate the forecast will become, constant revisions and
update is therefore needed. Furthermore, modelling based on some macro level assumptions
is also required (Malakooti, n.d.).
2.11 Identify demand pattern Time series data can reveal large variation of demand patterns. In order to ease the
forecasting process, it is beneficial to identify and classify those patterns. Moreover,
separating time series into its constituent components might also be used as a support tool to
purpose which forecasting technique to use in the process. In this chapter, we will reflect to
the most common data patterns and approaches used to extract the main components of the
data.
2.11.1 Demand patterns A time series can be defined as: “a set of observations ordered in time, on a given
phenomenon (target variable)” (Dagum, 2010). Time series data posses some special
features associated with the sequence of the observation, which is important in a time series.
In order to forecast, the authors have to make important assumptions such as, the
relationships between different variables affecting demand will continue in the future.
Newbold, Carlson and Thorne (2013) identify four main patterns of time series data that
forecasting is based on:
1. Trend pattern
2. Seasonality pattern
3. Cyclical pattern
4. Irregular pattern
These patterns are usually independent from one another, and together provide a
decomposition model that can be shown by the following equation (Dagum, 2010):
Xt = Tt + St + Ct + It
Where
Literature review
26
Xt denotes the observed series
Tt is the trend pattern
St is the seasonality pattern
Ct is the cyclical pattern
It is the irregular pattern
In the case where it exist some dependence among these patterns, the model can be shown in
multiplicative form (Newbold, Carlson & Thorne, 2013):
Xt = Tt St Ct It
The identification process is not always easy going. It is usual that the data does not fit to
any of above mentioned demand patterns, as the data may be influenced from several
directions at the same time (Jacobs, Chase and Lummus, 2011).
Trend pattern The term trend pattern is used for stable growth or decline of values over several successive
time periods and it can be described as a “long-term change in mean level per unit time”
(Chatfield, 2001). Trend patterns have two coefficients and their relation can be shown by
the equation:
μ(t) = a + bt
Where
a is the intercept on the y-axis
b is the slope of the trend curve (Thomopoulos, 2015)
There are a couple of forecasting models that are appropriate for forecasting of trend shaped
data, such as trend regression and trend smoothing forecasts (Thomopoulos, 2015). Jacobs,
Chase and Lummus (2011) describe four main types of trends:
1. Linear trend
2. S-curve trend
3. Asymptotic trend
4. Exponential trend
Literature review
27
A linear trend can be described as a straight continuous growth or decline of the demand
over long periods of time. A S-curve trend is typically used for description of a product
growth and its maturity cycle, with a main focus on the point of the curve where the trend
changes from slow growth to fast growth or vice versa. An asymmetric trend is
characterized with the highest demand growth at the beginning, which diminishes over time.
Contradictory to the asymmetric trend, an exponential trend indicates that the demand will
continue to grow exponentially over time (Jacobs, Chase and Lummus, 2011).
The identification of the long-term trend has been a serious challenge to statisticians due to
the fact that the trend is a non-observable component and can easily be mixed with long
business cycle. In order to avoid this kind of problem, statisticians have used couple of
solutions such as estimating the trend over the whole series (Dagum, 2010).
Seasonality pattern The seasonal pattern is associated with a period of the year characterized by some specific
activity (Jacobs, Chase and Lummus, 2011). This activity continues to repeat over the same
period of time for each year as a regular and oscillatory behaviour (Newbold, Carlson and
Thorne, 2013). Seasonality appears due to the fact that a period during the year is more
important in terms of activity than others. The most common causes of the seasonality are
weather, composition of the calendar, expectations and important institutional deadlines.
The identification of seasonality pattern starts with the recognition of at least one month or
quarter during the year, which tends to be more important in terms of activities or levels
(Dagum, 2010). Next step is to determine the seasonal factor or index, which can be
described as the amount of correction needed in a time series in order to adjust for the
season of the year. The seasonal factor should be updated as new data becomes available
(Jacobs, Chase and Lummus, 2011).
Cyclical pattern Apart from seasonal patterns, many businesses are characterized by oscillatory or cyclical
patterns that do not indicate annual repetitive activity (Newbold, Carlson & Thorne, 2013).
The duration of a cyclical pattern can last couple of years, however the length of the cycle is
unknown beforehand (Chatfield, 2001). It is common for a cyclical pattern to reach its peaks
during upswings and downswings of the economy. Cyclical pattern is characterized for
having fluctuations that are not of fixed period. The average lengths of cycles are much
longer than for seasonal pattern and the scale of a cycle tends to be very variable (Dagum,
Literature review
28
2010).
Irregular pattern In irregular pattern the data exhibits components that are unpredictable and irregular on the
basis of the past experience. This pattern can be identified both as a similar as well as
random error term in regression model (Newbold, Carlson & Thorne, 2013).
2.11.2 Decomposition of data Before computing a model for forecasting of certain data, it is suitable to check the data in
order to get some understanding of the data’s possible variations through time. The process
of identification and separation of data into its constituent components or data patterns is
called decomposition of the data. Using the decomposition of data process (Jacobs, Chase
and Lummus, 2011), together with tools such as the time plot, graphs or summary statistics,
the analyst will be able to identify and separate several patterns of demand for products or
services. The graphical analysis is valuable for describing the data, but it might also be
helpful as a support tool to propose models and theories for the forecasting process. These
graphical analysis tools should identify important characteristics such as trend, seasonality,
turning points or similar changes in the structure of the data (Chatfield, 2001). Software
such as SAS, Stata and R are available and easily implemented for interactive graphical
analysis (Lee Rodgers, Beasley and Sitchuelke, 2014). There are several ways to extract
associated components from a time series data. Some decomposition methods will be
explained based on a manual approach using mathematical calculations, while others will be
explained based on how to extract the main components of the data using software R.
The classical decomposition method The classical decomposition method is a quite simple procedure that forms the basis for
several other decomposition methods (Hyndman and Athanasopoulos, 2014). According to
Makridakis and Wheelwright (1987), the classical decomposition method consists of
following steps:
1. Estimate a moving average based on the length of seasonality; twelve terms
moving averages for yearly seasonality; three moving averages if the data is on the
quarterly basis
2. Provide seasonality ratios by dividing the actual data by the corresponding moving
average value
3. Calculate coefficients of seasonality by removing randomness from the seasonality
Literature review
29
ratios. This can be done by averaging all corresponding values (same period from
different years)
4. In order to extract the seasonality from the data, the original data should be
divided by the coefficient of seasonality. When this is done, the data will still
include the other three patterns: trend, cycle and randomness
5. In order to remove randomness, it is necessary to compute three or five moving
averages terms of the deseasonalized data (Makridakis and Wheelwright, 1987)
Two additional steps could be used in order to improve the curve of the trend-cycle
component that was obtained in step five:
. a) When the number of average terms is even, the analyst should centre the moving
average by putting it in the middle of the averaged N data values in step 1. When the
length of the seasonality is odd, the average is directly cantered (Makridakis and
Wheelwright, 1987)
. b) In step three a medial average should be used, which is done by eliminating the highest
and the lowest value before averaging the ratio of seasonality (Makridakis and
Wheelwright, 1987)
Extraction of the seasonal component through moving averages Analysts may want to remove seasonal components from the series in order to obtain a
brighter appreciation of other data components behaviour. As data is very often presented as
a quarterly time series, producing four-period moving averages can help to remove
seasonality. This means that various seasonal values are brought together in a single
seasonal moving average (Newbold, Carlson & Thorne, 2013). Computing the first member
of the series is shown in the following equation:
X*
2.5 = (X1 + X2 + X3 + X4) / 4
Where
X*
2.5 = four point moving average
X1 = the value for the quarter 1 (Newbold, Carlson & Thorne, 2013).
Literature review
30
Computing the second member of the series is shown in equation bellow:
X*
3.5 = (X2 + X3 + X4 + X5) / 4
Where X*
3.5 = is four point moving average
X2 = is the value for the quarter 2 (Newbold, Carlson & Thorne, 2013).
The new series of moving averages still have an issue, even though they are free from
seasonality. The location of the members of the moving averages series does not correspond
with that of the initial series, as the first term is the average of the first four values or i.e. it is
being centred between the second and third observation. The same issue occurs to the other
terms of the series. In order to solve this problem, one has to centre the series of four-point
moving averages. Centring the series of four-point moving averages can be achieved by
estimating the averages of nearby pairs (Newbold, Carlson & Thorne, 2013).
X*
3 = (X*
2.5 + X*
3.5) / 2
Where
X*
3 = the centred moving average corresponding to the third observation of the initial series
Using this method, the analyst is able to remove completely the seasonal as well as to
smooth the irregular component. The final result of this procedure is an ability to judge the
non-seasonal regularities of the data (Newbold, Carlson & Thorne, 2013).
Decomposition of the data using software R R is widely used software aimed for manipulation, calculation and graphical display of the
data providing an effective data handling and analysis. It was developed rapidly using
extensive packages assisting easy adaption for newly developing methods of interactive data
analysis (Venables and Smith, 2009). The main advantages of this software are that it is free,
maintained by scientist for scientists and available for every operative system. It is very
useful for decomposition of time series data, making the procedures simple compared to the
previously explained decomposition processes that are computed on the manually basis
(Zucchini and Nenadic, n.d).
Literature review
31
The process starts by importing the data into software R that may be done in different ways,
mostly as values from external files rather than entering during an R session (Venables and
Smith, 2009). Various codes are used for different decomposition methods in R. In order to
compute the classical decomposition, following code should be used:
Fit ← decompose (x, type= “multiplicative”)
Where
Fit is the result of the decomposition
X is the name of the time series that has been imported into R
“Multiplicative” is type of time series, it can also be additive (Hyndman and Athanasopoulos, 2015)
Moreover, evaluation of trend in a time series data can be done using nonparametric
regression techniques. Using the function stl ( ), the software R will perform a seasonal
decomposition of a given time series by determining the trend Tt using “loess” regression.
Furthermore, the software calculates the seasonal component (Zucchini and Nenadic, n.d).
By using software R and its codes, it is possible to perform different decomposition methods
as well as forecasting procedures. For example, using the function HoltWinters (x, alpha,
beta, gamma), the analyst can specify the three smoothing parameters by himself / herself.
The analyst can also avoid specifying the particular components alpha, beta and gamma. In
that case, the software will determine smoothing parameters automatically and compute the
forecasts (Zucchini and Nenadic, n.d). More detailed information on how to use the software
and its functions is provided in the program browser and it is available offline (Reiner,
2014).
2.12 Choose forecasting technique The implementation of forecasting to produce numerical estimates ranges from relatively
simple techniques to complex methods (Jarrett, 1987). The selection of a method depends on
the context of the forecast, availability of historical data, degree of accuracy desirable, time
period to be forecast and time availability for making the analysis. These elements must
continuously be weighed and a technique that provides the greatest benefits for the company
should be chosen. Forecasting techniques can broadly be categorized into quantitative,
qualitative and causal methods depending upon the extent to which mathematical and
statistical methods are used (Chambers, Mullick and Smith, 1971).
Literature review
32
Most techniques are quantitative in nature, which means that aspects of the world are
translated into mathematical analysis (Jarrett, 1987). The quantitative approach focuses on
statistical analysis on past demand to generate forecasts. A basic assumption is that the
underlying trend of the past will continue into the future. Qualitative forecasting techniques
generally employ the judgment of experts to generate forecasts. This technique is often
implemented when numeric data is not available, or when data is not interpretable by
quantitative means alone. Causal forecasting methods are based on cause-and-effect
relationship between the variable to be forecasted and an independent variable. The causal
method seeks to establish direct relationships between demand and factors influencing it. By
analysing past data, one can forecast future demand of an item (Thomopoulos, 1980).
2.12.1 Quantitative methods
2.12.1.1 Simple moving average The simple moving average is the most common forecasting technique used. The use of the
moving average involves calculating the average of a sample observation and employs that
average as the forecast for the next period. Moving averages are lagging indicators, an
economic factor that changes once the economy follows a particular trend. Consequently
moving average do not predict trends, but rather confirm their current direction and progress
(Cortinhas and Black, 2012). As each new sample observation becomes available, a new
average is calculated by removing the latest sample observation from the average and
includes the most recent figures. Thus, each forecast is recomputed as new data becomes
accessible.
The algebraic formula for the moving average:
𝐹𝐹 =𝐷𝐷𝐷𝐷𝐷𝐷 𝑖𝐷 𝑝𝑝𝐷𝑝𝑖𝑝𝑝𝑝 𝐷 𝑝𝐷𝑝𝑖𝑝𝐷𝑝
𝐷
A critical element in calculating the moving average is the number of time periods used. It is
central to find a moving average that will be consistently profitable. The length of a moving
average should fit the market cycle one wish to follow (Cortinhas and black, 2012). The
benefit of the moving average is its simplicity. The technique is easy to understand, compute
and it provide stable forecasts. However, since it’s a trend following model it only works
well when the market is trending. In a stable market the lags of the market will cause the
moving average to generate false signals (Jarrett, 1987).
Literature review
33
At occasions a forecaster may want to assign more weight on certain period in time, since
the value from one month can be considered more relevant than data from previous months.
By assigning more weight to each value in the data series according to its age, the most
recent data gets the greatest weight and each previous data value gets a smaller weight as
one move backward in the series. This is known as weighted moving average. The weighted
moving is also a lagging indicator that emphasizes the direction of a trend and smooth out
price and volume fluctuations. However the weighted moving average differs from the
simple moving average that weighs all periods equally. Since the weighted moving average
assigns more importance to recent values, it is more sensitive to trends activity than the
simple moving average (Taylor, 2008).
The algebraic formula for the weighted moving average:
𝐹𝐹 + 1 =𝛴 (𝑊𝐷𝑖𝑊ℎ𝐹 𝑓𝑝𝑝 𝑝𝐷𝑝𝑖𝑝𝐷 𝐷)(𝐷𝐷𝐷𝐷𝐷𝐷 𝑖 𝑝𝐷𝑝𝑖𝑝𝐷 𝐷)
𝛴 𝑊𝐷𝑖𝑊ℎ𝐹𝑝
2.12.1.2 Exponential smoothing Another method that is very similar to the moving average method, in form of forecasting
results, is exponential smoothing. Exponential smoothing method is applied for computing
forecasts with simple updating formula (Axsäter, 2006). According to Chatfield (2001),
exponential smoothing method should be used when there is large number of series to
forecast and analyst`s skills are limited as the method is relatively simple and automatic.
This method can only be applied to already known values from the past in order to predict
future series values that are unknown (Snyder, 2006).
Billah et al. (2006) state that there exist several variations of this method that are able to
follow changes on various levels such as trends and seasonality. Based on what one want to
include in the forecasting model, one can distinguish these three basic variations of
exponential smoothing forecasting method:
1. Simple exponential smoothing
2. Holt-Winters non-seasonal method
3. Holt-Winters seasonal method
Two main characteristics of these methods is that time series are computed from
components such as level, growth and seasonal effects and that these components need to be
Literature review
34
updated over time (Billah et al., 2006).
Simple exponential smoothing The simple exponential smoothing model (SES) is based on an assumption that forecast data
should fluctuate around a constant level or changes slightly over the time (Ostertagová and
Ostertag, 2012).
The SES model can be described by the model equation:
𝐹𝐹 + 1 = 𝛼 𝑌𝐹 + (1 − 𝛼)𝐹𝐹
Where
Ft + 1 = the new forecast
Yt = actual series value at the time t
Ft = forecast value of the actual series value at the time t
α = Smoothing constant (0 < α < 1)
According to Ostertagová and Ostertag (2012), the new forecast Ft+1 is based on weighting
the latest observation or actual series value Yt with weight α and weighting the latest
forecast value of the actual series value Ft with a weight 1 - α. Smoothing constant is
subjectively chosen and ranges from 0 to 1. It is also possible to choose α = 0, which means
that one doesn’t update the forecast and α = 1, which means that one uses the latest series
value as ones forecast (Axsäter, 2006).
The accuracy of this forecasting method depends on the smoothing constant α. When
choosing an appropriate smoothing constant, the forecasting error will be minimized. It is
essential to choose a smoothing constant that balances the benefits of smoothing random
variations, while being able to respond to real changes if they occur. When α is close to 1,
the new forecast will be substantially different from the previous one and it is useful when
the underlying average is likely to change. Low values of α are used when underlying
average is stabile, which results in very similar forecast as the previous one (Ostertagová
and Ostertag, 2012).
Holt-Winters non-seasonal method Trend-corrected exponential smoothing has shown to be accurate in various empirical
Literature review
35
studies over the last twenty years (McKenzie and Gardner, 2010). This forecasting method
includes a trend term Tt that is supposed to measure expected increase or decrease per unit
time period in the local mean level (Chatfield, 2001). The biggest difference between this
model and the simple exponential smoothing model is that forecasts for future periods are no
longer the same, as trend or change per period can be negative (Axsäter, 2006). According to
Middel (2014), the Holt- Winters non- seasonal method can be explained by the equation:
FITt = Xt + Tt
Where
FITt = Forecast including trend
Xt = Exponential smoothed forecast
Tt = Exponential smoothed trend
Xt = (1-α) (Xt-1 + Tt-1) + α Yt
Tt = (1-ß) Tt-1 + ß (Xt - Xt-1)
α & ß are smoothing constants between 0 and 1.
Holt-Winters seasonal method Holt-Winters seasonal method is used when a seasonal pattern exists in the data. The method
is conducted by extending the previously mentioned non- seasonal method with a smoothed
seasonal factor Ft for each period of the year. This factor is used in order to adjust the
forecasts according to the expected seasonal fluctuations (Makridakis and Wheelwright,
1987). As in non- seasonal method, Yt, Xt and Tt represent the observed value, the level and
trend estimates at time t. If time series contain s periods per year, the seasonal factor for the
corresponding period in the year before will be Ft-s (Axsäter, 2006). The estimates of level,
trend and seasonal adjustments are updated by following equations:
Xt = (1- α) (Xt-1 + Tt-1) + α (Yt / Ft-s)
Tt = (1-ß) Tt-1 + ß (Xt - Xt-1)
Ft = (1-ɣ) Ft-s + (Yt / Xt)
Where
Literature review
36
Xt is the smoothed level of the series
Tt is the smoothed trend of the series
Ft is the smoothed seasonal adjustment for the series
Yt is the observed value at the time t
α is the smoothing level constant
ß is the smoothing trend constant
ɣ is the smoothing multiplicative seasonal constant
s is a period of the time series (Newbold, Carlson & Thorne, 2013)
The term (Xt-1 + Tt-1) is used for estimation of the level at time t computed at the previous
time period t-1. Xt is then updated when Yt is available, removing the influence of the
seasonality by deflating it by the latest estimate Ft-s. The updating equation for trend is the
same as for the non- seasonal model. Regarding the seasonal adjustment Ft, the most recent
estimate of the factor is from the previous year Ft-s. In order to complete the seasonal
adjustment, the new observation Yt is divided by the level estimate Xt. These equations are
used for forecasting future volumes at h time periods from the last observation Yn in the
historical series (Newbold, Carlson & Thorne, 2013). The final algebraic equation:
Xn+h = (Xn + hTn) Fn + h-s
The Holt-Winters seasonal method uses three smoothing parameters to update level, trend
and seasonal component at each period. The procedure explained above is based on the
multiplicative option and the actual forecasts will depend on the values chosen for the
smoothing parameters. The choice of the smoothing parameters can be based on either
subjective or objective criteria. A possible alternative could be to test several different sets
of possible smoothing parameter values on the available historical data. The set that
provides the best forecasts for the data should be used to generate future forecasts (Newbold,
Carlson & Thorne, 2013).
Literature review
37
2.12.2 Casual methods
2.12.2.1 Simple regression analysis Regression analysis is a mathematical model used to estimate the nature between a
dependent variable and an independent variable. The technique studies historical
relationship between variables in order to predict their future value behaviour. Regression
models can be either linear or nonlinear and simple or multiple (Ambrosius, 2007). For the
sake of applicability only linear simple regression analysis will be concerned with. A linear
model graphs the relationships between variables as a straight line, while a nonlinear model
illustrates the relationships between variables as curved lines (Jarrett, 1987). The most
fundamental regression model is the simple regression, where two variables are assumed to
be systematically connected by a linear correlation. By studying the correlation between the
dependent variable (Y), the variable whose value is to be predicted and the independent
variable (X), the degree of linear association between the two variables can be determined
(Shalabh, 2013).
The formula for simple regression line:
y= a + bx
Where
y = Predicted (dependent) variable
x = Predictor (independent) variable
b = Slope of the line
a = Value of y when x = 0
When conducting a simple regression analysis a few assumptions need to be made in order
to create a strong linear relationship between dependent and independent variables:
First assumption: the relationship between X and Y is linear.
Second assumption: variable x increases as variable y decreases.
Third assumption: Random error (e) is assumed to follow the normal distribution, E (e) = 0 (Thomopoulos, 1980).
Literature review
38
The first step in regression analysis is to collect historical data regarding the two variables in
order to construct a scatter plot. By creating a scatter plot (in excel) one can graph the data
to yield information regarding the spread of the figures. This allows the user to examine if
the regression line fits through the points of data. Typically the regression line doesn’t pass
through all of the points, the vertical distance between each point and the line is the error of
the prediction. The coefficients a and b of the line are based on the following two equations:
𝑏 =𝐷( 𝛴𝛴𝛴)− (𝛴𝛴)(𝛴𝛴)𝐷(𝛴𝛴2) − (𝛴𝛴)2
𝐷 = 𝛴𝛴 − 𝑏𝛴𝛴
𝐷
Where
n = Number of paired observations
Once the straight line is determined one will see points that are outside the straight line,
these are prediction. This can be caused due to various reasons, for example incorrect
measurement of data. If strong statistical evidence against the outlier is presented, the outlier
should be corrected if possible otherwise it needs to be deleted from the data set
(Ambrosius, 2007). The strength and direction of the linear relationship between the
variables can be measured by calculating the correlation coefficient (r). The correlation
coefficient is a value between -1 and +1. If the calculation of r is proven to be high, the
relationship between the variables is strong. Conversely, if the value of r is proven to be
negative, one will experience a negative correlation (Cortinhas and black, 2012).
The formula for correlation coefficient is:
𝑝 = 𝐷𝛴(𝛴𝛴) − (𝛴𝛴)(𝛴𝛴)
�𝐷(𝛴(𝛴)2 − (𝛴𝛴)2 �𝐷(𝛴𝛴2) − (𝛴𝛴)2
The square of the correlation coefficient r2 indicates how much the depended variable can be
explained from the independent variable. In order to generate an accurate forecast a variable
with a maximum correlation needs to be identified, i.e. a correlation coefficient with ±1 is
required (Ambrosius, 2007).
Literature review
39
2.12.3 Benefits and drawbacks of different forecasting techniques The primary goal of a forecasting technique is to provide valuable information to a business
in order to ease future decision. A corporation can employ a diverse range of forecasting
techniques to generate potential future results. Each technique has a set of advantages and
disadvantages, which needs to be carefully weighted. The most notable advantage of
quantitative forecast techniques is that the projection of the future relies on the strength of
past data (Makridakis and Wheelwright, 1987). Quantitative forecasts such as moving
average and weighted moving average are easy to compute, implement and understand. Still,
these methods have some drawbacks. If a method is considered too easy it can cause
thoughtlessness, particularly in a long- term time horizon. Since both set of methods are
trend following models they works well when the market is trending. In a stable market
however the lags of the market will cause the techniques to generate false signals (Jarrett,
1987).
An additional qualitative forecasting technique is exponential smoothing. The benefits of the
technique are several. By using exponential smoothing one can easily computerize a large
number of a variable. There is no limit to how many parameters to estimate, thus making it
easy to update the model if a new realization becomes available (Brooks, 2002). Moreover,
the technique enables the user to set up monitoring schemes that are easy to understand.
Unlike moving average, exponential smoothing does not lag behind actual trend and can
therefore produce more accurate forecasts. The disadvantage exposed to the model is that it
can fail to produce accurate forecasts, as finding the best smoothing constant can be
problematic (Makridakis and Wheelwright, 1987). Long- term forecasts can be overly
affected by events in the series under investigation (Brooks, 2002).
The principal advantage of causal forecasting techniques such as simple and multiple
regression models is simplicity and the ability to see interactive patterns. The result obtained
from regression models provides a link, that is to say a correlation, between various
variables. The techniques use data efficiently, as good results can be obtained with relatively
small data set (Harrell, 2001). Consequently there are a few drawbacks with regression
models. The main disadvantage is the fact that regression models only determine
relationship between different variables, but never explain the underlying reason for the
correlation. Regression models are also exposed to parameter instability, as there is a
tendency for relationship between variables to change over time due to external changes.
Literature review
40
Another drawback of the techniques is their sensitivity to outliers (Makridakis and
Wheelwright, 1987). Outliers represent error of the prediction and must be considered when
computing a regression model (Ambrosius, 2007).
2.13 Evaluation of forecasting technique and result Choosing the best forecasting technique is not a simple task as it depends on what is meant
by the best. Some methods that are appropriate and successfully used by one forecaster may
be inappropriate for another forecaster. Choosing the best does not always mean choosing
the most accurate method, while ignoring other factors. A simple method that is marginally
less accurate than a more complex one could be preferred in practice (Chatfield, 2001).
2.13.1 Forecast accuracy, also known as error measurement The accuracy of a forecast is the degree of freedom from error. Forecast error measurement
plays a critical role in tracking forecast accuracy. Level of forecast accuracy can enhance
business effectiveness to serve demand while lowering operational cost. Forecast error
measurement has become a critical part of the forecasting process, as it enables companies
to evaluate the success of chosen forecasting technique. In simple, forecast error is the
difference between the actual value that has occurred and the value that was predicted for a
given time period (Stevenson, 2005). Various statistical measures can be used to measure
the accuracy of a technique, or in other words forecast error. The most common statistical
error measurements for summarizing historical errors include mean absolute deviation
(MAD), mean square error (MSE) and mean absolute per cent error (MAPE) (Shim and
Siegel, 1988). Any of several methods can be used to compute forecasting error. Selection of
a specific error measurement is determined by the objectives of the forecaster and the
forecasting process (Cortinhas and Black, 2012).
Mean Absolute Deviation (MAD) The mean absolute deviation is defined as:
𝑀𝑀𝐷 = 𝛴 |𝑀𝐹 − 𝐹𝐹|
𝐷
Where
At = Actual demand for period t
Ft = Forecast for period t
n = number of periods of evaluation
Literature review
41
MAD is the average absolute error and measures the absolute value of forecast error and
averages the error over the entity of the forecast time periods. An examination of forecast
data can reveal both positive and negative forecast errors. By summing these errors in order
to measure an overall measure of error, the positive and negative values can offset each
other, thus resulting in an underestimation of the total error. MAD resolves this issue by
taking the absolute value of error measurement. Thus analysing the magnitude of forecast
error with no concern to direction (Cortinhas and Black, 2012). The smaller the value of
MAD the more accurate the forecast is (Stevenson, 2005).
Mean Square Error (MSE) The mean square error is defined as:
𝑀𝑀𝑀 = 𝛴(𝑀𝐹 − 𝐹𝐹)2
𝐷 − 1
Where
At = Actual demand for Period t
Ft = Forecast at period t
n = number of periods of evaluation
MSE calculates the mean square error. Such squaring places substantially more weight to
large errors than smaller ones. The smaller the MSE value the more stable the forecasting
technique is (Makridakis, 1995).
Mean Absolute Percentage Error (MAPE) Mean absolute percentage error is defined as:
𝑀𝑀𝑀𝑀 = 𝛴|𝑀𝐹 − 𝐹𝐹|
|𝑀𝐹|𝛴 100
Where
At = Actual demand for Period t
Ft = Forecast at period t
n = number of periods of evaluation
Literature review
42
MAPE expresses error as a percentage of actual data, as it measures the size of error in
percentage terms. The measurement is easy to interpret, however it is very scale sensitive
and shouldn’t be implemented when working with low volume data (Makridakis, 1995).
2.13.2 Evaluation of forecasting results Danese and Kalchschmidt (2011) highlight the importance of information flow when
conducting a forecasting process, which means that the right information has to be shared to
the right source at the right time. In order to achieve this, forecasting roles have to be
assigned to each department so that everybody clearly understands their part in the process.
Furthermore, judgment abilities play a crucial role in the process of evaluation. Thus, to
understand the process of evaluation it is important to specify in which context the
evaluation process occurs and what affects the actual outcome (Makridakis and
Wheelwright, 1987). Hogarth (1980; Makridakis and Wheelwright, 1987) computes a
conceptual model of judgment involving three elements:
1. Person
2. The task environment, where the person is able to make judgments
3. Actions that result from judgments. These actions can affect both the person and the
task environment.
In this model, one`s personal belief and value system related to the judgmental task is
represented by his or her “schema”, while operations of judgment consist of:
1. Acquisition of information aimed for judgment
2. Processing of the gathered information
3. Output of the judgmental process (Makridakis and Wheelwright, 1987).
Furthermore, the output of the judgmental process implies actions. These actions together
with external factors yield outcome that sends feedback affecting person`s value system and
the task environment. At the same time, the person`s value systems affect complexity of the
environment, the problem identification, actions taken and their objectives. Links between
person`s schema, actions, outcomes and feedback to the person`s schema are essential. If
these links are short (in time) and frequent then the person can easily understand and
interpret the effects of actions on outcomes. This means that if the action-outcome-feedback
link is neither short nor clear, the person will not be able to improve its decision making
Literature review
43
process (Makridakis and Wheelwright, 1987). According to Schultz (1992), the final
outcome shows if the organization is better off by having changed the forecasting method or
process. This step goes beyond measuring the accuracy of the forecasting method. Instead, it
looks at the objective performance measures such as sales, cost and ultimately profit, as that
is what forecasting in organizations is all about. The conceptual model of judgment is shown
in the following picture:
Figure 7. Conceptual model of judgment (Makridakis and Wheelwright, 1987)
2.14 Integration of forecasting process Forecast integration across an entire organization is an important aspect of the forecasting
process (Makridakis and Wheelwright, 1987), which can lead to better coordination between
departments as well as its alignment to the corporate strategy (Pal Singh Toor and Dhir,
2011). Makridakis and Wheelwright (1987) differ between two types of forecasts, implicit
Literature review
44
and explicit. The major feature of implicit forecasts is that these forecasts are not
incorporated into plans and decisions being made. On the other hand, explicit forecasts are
used directly for planning and decision-making purposes. Organizations may not always be
interested in improving forecast accuracy, as this isn't always considered as a priority.
However, organizations can achieve significant improvements regarding their performance
through a good forecasting management and its integration (Danese and Kalchschmidt,
2011).
Integrating the forecasting process refers to connecting technologies, applications and
processes across the entire organization in order to enhance transparency, organizational
alignment and finally financial performance. Once this has been done, a well-integrated
forecasting process will link strategic targets with tactical and operative planning. This
integration process may successfully be used in a number of areas such as capital
expenditures, supply chain relations and other business processes that may affect the
competitive strategy of the company. The benefits of an integrated forecasting process are:
• Ease evaluation process based on the true economic impact
• Improve strategic alignment and position
• Increase transparency
• Provide decision support
• Optimize asset utilization
• Improve financial performance (Pal Singh Toor and Dhir, 2011).
Integration, monitoring and control of forecasting process are a complex task, as it requires
managerial skills combined with the ability to handle politics within organization. The high
level of complexity enhances importance of the integration process, as good forecasts can
sometimes fail through poor planning and their implementation (Makridakis and
Wheelwright, 1987). Monitoring forecasts in conjunction with the evaluation of forecasting
errors is a vital part of the on-going forecasting process. These actions need to be
implemented in order to accomplish continuous improvements and enable adjustments
accordingly (Ord and Fildes, 2013).
Literature review
45
Figure 8. The forecasting system (Makridakis and Wheelwright, 1987). How forecasts can be integrated into a decision making process can be seen in figure 8. The
highlights of the model are in the forecaster`s and decision maker`s section as well as in
planning guidelines & assumptions and forecast recommendations. In this model the
forecaster provides the decision maker with the information necessary to reach an action
plan. The forecaster initiates a forecasting procedure based on planning guidelines and
assumptions that include constraints of costs, time, accuracy of the methods he or she is
competent to perform and finally his or hers own values. The forecaster and the decision
maker have two different sets of values that may not match. That means that the forecaster
may be influenced by decision maker`s values. On the other hand, the decision maker is
influenced by forecast recommendation, but he or she may modify it using alternative
sources of information. If the working forecast is satisfactory then the action plan is
computed based on the forecast recommendations, otherwise, the planning assumptions are
revised and sent back to the forecaster (Makridakis and Wheelwright, 1987).
Literature review
46
2.15 Summary of the literature review In order to easily identify important factors in each step of the forecasting process a summary has been defined. By using key words one can easily understand the essential factors of each step. Hence, a simple overview of the forecasting steps can be seen below.
Figure 9. A summary of the literature review
IDENTIFY STRATEGIC GOALS
Strategic goal alignment Identification of KPIs
DECIDE OBJECTIVES OF THE FORECAST Level of sophistication Trade -off between costs and the level of accuracy
CHOOSE A FORECAST APPROACH Exploring historical data Exploring other factors that could affect forecasting process
CHOOSE FORECAST VARIABLE FROM COLLECTED DATA A proper selection of variables Clean the data
CHOOSE FORECAST HORIZON Short, medium and long-term forecasting
IDENTIFY DEMAND PATTERN Trend, seasonal, cyclical and irregular patterns Decomposition of the data
CHOOSE FORECAST TECHNIQUE Moving average, exponential smoothing, linear regression etc
EVALUATION OF FORECASTING TECHNIQUES AND RESULTS MAD, MSE & MAPE Process of evaluation of forecasting results
INTEGRATION OF FORECAST PROCESS Incorporation of forecasts into future decisions through forecast recommendations and Monitoring results
Methodology
3. Methodology The aim of this chapter is to give a detailed presentation of employed methods that has been conducted to reach the final outcome of the thesis. An explanation of the research approach, strategy and design has been provided. Followed by a classification of the research based on logic of the research. Furthermore, methods of data collection are presented followed by a research evaluation.
3.1 Research approach Recognizing the significance of the forecasting process as a mean of sustaining competitive
advantages was the initial stage of the research process. It became obvious that forecasting is
a complex discipline and that its alignment with the corporate strategy of the company
would improve the overall financial performance. Thus, in order to get a better impression
regarding the concept of forecasting as a process, the authors conducted a literature review
that helped to recognize and understand the main steps of a forecasting process. During the
progress the authors realized quickly that no single theory explains all the important
elements of the forecasting procedure and its alignment with corporate strategy of a
company.
Hence, the authors developed a theoretical framework consisting of theories connected to
the forecasting process. By working on this theoretical framework, the authors were able to
compute a model that could serve as a basis to statistically and theoretically evaluate the
process of forecasting and its accuracy. Thus outline how companies align this process with
strategic goals and integrate it across an organization in order to obtain competitive
advantages. Once the empirical findings were gathered, the authors compared the existing
forecasting model of the case company with the conceptual forecasting model. As these
models showed huge differences, the authors suggested the conceptual model as a useful
forecasting process for the case company.
3.2 Research design A research design defines the overall strategy of the thesis. It’s a strategy to help guide a
research towards its purpose. A research design provides a framework for collection and
analysis of data. The design of the process involves several decisions about the way that the
research will be carried out. One type of research design is the case study design. A case
Methodology
48
study is a detailed analysis of a restricted number of events and their interrelations (Bryman
and Bell, 2003). The design chosen for this master thesis is a case study. The aim is not
merely to explore the forecasting discipline, but rather to create a detailed and in- depth
understanding of the subject, as well as to test scientific theories and models and their
usefulness in real world environments. The implementation of a case study is therefore a
suitable approach for understanding the complex situation of the case company.
3.3 Research strategy- mixed methods The mixed method strategy is a process of using both quantitative and qualitative research
methods in a single study in order to investigate a research problem. The strategy is used
when a quantitative and qualitative research strategy together provides a better
understanding of the research problem then either type separately. The mixed method design
is based on the incorporation of qualitative components in an otherwise quantitative study.
Qualitative and quantitative instruments process data in order to obtain more detailed
information (Bryman and Bell, 2003). According to Teddlie and Tashakkori, (2009) there
are three approaches to mixed method research: triangulation, facilitation and
complementarity.
Triangulation is the combination of methodologies in a single study. It’s the concept of cross
checking the results generated from one research strategy with the results generated from
another research strategy. By employing a qualitative research in relation to a quantitative,
one can increase the confidence in the findings derived from a quantitative research strategy.
Facilitation is the use of one data collection method to aid another data collection method.
Facilitation can be done in two ways, either by facilitating a quantitative research with the
use of a qualitative research or quantitative research can facilitate a qualitative research
(Bryman and Bell, 2003). There are several ways in which qualitative research can guide a
quantitative research study. One way is by using qualitative research as a source of
hypothesis, which subsequently can be tested using quantitative methods. Another way is to
form the design of survey question through qualitative instruments, than use quantitative
research to measure the outcomes. While, quantitative research is used to formulate the basis
for qualitative research through the selection of participants for interviews (Saunders, Lewis
and Thornhill, 2003). The complementarity approach is used when a researcher can’t rely
merely on a quantitative or qualitative method alone, and must fill in the gaps with the
support of another research strategy (Bryman and Bell, 2003).
Methodology
49
Due to the nature of this research a mixed method strategy is considered appropriate. By
implementing both qualitative and quantitative instruments one will reduce possible threats
that each instruments faces individually. The strategy has received much criticism from
different researchers, claiming that mixed methods are not possible due to the
incompatibility of the paradigms underlying them (Bryman, 2007). However, the researchers
of this study believes that by integrating both research strategies, rather than keeping them
separate, one will be able to maximize the strength of the qualitative and quantitative data.
By merging data from the two strategies one is able to develop a more complete
understanding of the problem. By applying triangulation one doesn’t only rely merely on
one research strategy rather one is free to gather and interpret information, in relation to
numeric data, in order to create an in depth understanding of the problem. Through the
implementation of a mixed method strategy one is able to combine techniques found in both
qualitative and quantitative research method and address both numerical, as well as
exploratory questions. The mixed method approach will enable the researchers to conduct
both a statistical and a textual analysis, thus create a conclusion based upon both qualitative
and qualitative perspectives.
The quantitative aspect of the study aimed at identifying variables that influence container
volumes in combination with collecting empirical data provided from company X. The
purpose was to insert these figures in the presented forecasting techniques, in order to
compare and evaluate produced results. These were later compared through different
forecasting error measurements. By comparing statistical data, the researchers were able to
identify which forecasting technique that produced the greatest level of accuracy. Whereas
the qualitative components of the research study aimed at comparing different forecasting
processes, explaining the various elements forming the process, to ultimately compose an
improved forecasting process for prediction of container volumes for company X.
3.4 Research ontology- Interpretivism and positivism The positivist approach believes that the world is objective and researchers can stay
independent of the subject matter. The investigator and investigated are independent entities,
where the researcher is able of studying the phenomenon without influencing it or being
influenced by it (Guba and Lincoln, 1994). The positivist approach attempts to explain
causes of phenomena by using objective theories. The goal is to measure and analyze the
relationships between variables without manipulating them. Researchers of positivism
Methodology
50
studies need to remain detached of the subject matter. They need to preserve a distance
between themselves and the research object, in order to keep a clear distinction between
facts and value judgment. Statistical tools for quantitative processing of data are essential in
the positivist approach (Carson, 2001). This is because quantitative research only pursues
precise measurements (numerical evidence) to find the answer to an investigation (Collis
and Hussey, 2003).
The interpretivism ontology on the other hand believes that reality is subjective. The
ontology refers to how researchers make sense of their subjective reality and attach meaning
to it. The researcher and the object of study are interactively linked and findings are thereof
mutually created. The goal of interpretivist philosophy is to understand the meaning of
human behaviour. As a result the ontology relies heavily on naturalistic methods. The
structural framework of this ontology is more flexible than a positivistic research. The key to
the interpretivist paradigm is the acknowledgment that researchers cannot avoid to affect the
phenomena they study (Collis and Hussey, 2003).
Researchers of the mixed method approach don’t place much attachment to either ontology.
They believe that a mixed method research has neither an interpretivism approach nor
positivism (Piquero and Weisburd, 2010). According to Guba and Lincoln (1994)
methodological choices should be based on the research question and not on ideological
commitments to a paradigm. According to Howe (1992) there is no “either-or” decision to
be made, rather researchers of mixed method studies should embrace interpretivism with a
degree of positivism. Many researchers suggest that multiple paradigms can be incorporated
into a mixed method study, but each paradigm is related to a different phase of the research
study (Denzin and Lincoln, 2005).
The researchers of this study believe that compatibility can be found between positivism and
the understanding of interpretivism. The aim for positivist analysis is to come as close as
possible to the objective truth about the causes of events without influencing the subject of
study. This can be accomplished through statistical calculation of different forecasting
techniques and measurements of error. The objective for the interpretivist approach is to
understand how individuals understand their context and how the multiple subjective truths
provide insight into their corporate understandings. By using an interpretivist approach to
address questions regarding construction and accuracy of forecasting processes, one can
Methodology
51
understand the reality of the situation and make a subjective interpretation of the
phenomena. By combining the two paradigms one can generate greater analytical value.
3.5 Classification of the research based on logic of the research A deductive approach aims at developing a hypothesis based on existing theories, in order to
design a research strategy to test the hypotheses (Gratton and Jones, 2004). Deductive
reasoning is known to proceed from general principles to a specific conclusion, where a
particular theory is tested in order to determine its validity in given situations. By
formulating a set of hypothesis and applying relevant methodologies, a researcher can
confirm or reject a hypothesis (Bryman and Bell, 2003). An inductive approach on the other
hand, follows the logic of proceeding from empirical findings to theoretical results, one
move from specific observations to broader generalizations and theories. In an inductive
research, related theories are not given much attention until much further along the study.
Theories evolve as a result of the research. It is worth mentioning that most research studies
include both an inductive and deductive method some time during the research process
(Gratton and Jones, 2004).
In a mixed method research a deductive approach is applied to elements that show strong
theoretical ties, while an inductive approach is applied to elements that require an
exploratory investigation (Greene, 2007). In an inductive- deductive research one moves
from a set of observations to general conclusions through an inductive approach, then from
those general conclusions, one applies the deductive approach to test the hypothesis. Which
approach that is applied in the initial phase depends on where one is in terms of studying the
phenomena (Teddlie and Tashakkori, 2009). By implementing both research approaches one
is able to search for linearity between variables, while exploring empirical findings in order
to seek theoretical evidence to support the conclusion. Since the aim of this study is to test
differences between forecasting techniques and statistical errors while explaining and
uncovering the phenomena of forecasting as a process, an implementation of both
approaches is considered appropriate.
3.6 Method of data collection 3.6.1 Primary data Primary data is data collected for a specific research purpose. It means that data has been
collected from an original source first hand. The advantage of using primary data is that
Methodology
52
information is collected for the sole purposes of ones study. Questions are tailored
specifically to help a researcher extract data to help them with their investigation. The use of
primary data ensures the researchers that data gathered are meaningful to the purpose.
Secondary data lack the advantage of primary data, as secondary data is information
collected for another purpose (Walliman, 2011). Primary data for this research has been
collected through interviews with personnel working within different departments of
company X. The researchers of this study consider interviews to be an optimal way of
achieving efficiency in data collection.
3.6.2 Interview participants When deciding upon interview participants the aim was to identify respondents with most
relevant knowledge, as different respondent can describe the same issue in different way due
to their position, work and knowledge. Four interviews have been conducted with four
departments within company X. The first interview took place 25.03.2015 with the logistics
department. The participants were operations and logistics manager as well as logistics
coordinators. This interview created the basis for the empirical findings, as the participants
are the ones whom conduct the forecasting process. The next interview took place in
01.04.2015 with reporting manager. The reporting manager explained the ways of the export
department and their collaboration with the logistics department. Two additional interviews
were conducted in 15.04.2015 with trade import manager and import assistant and in
22.04.2015 with general commercial manager export. These additional interviews created a
holistic understanding of how all departments working separately and together as one unit.
3.6.3 Interview method and structure All interviews have been performed in a semi- structured nature. A semi-structured
interview combines predetermined set of open questions with the opportunity for the
interviewer to elaborate particular responses further (Yin, 2009). The benefits of a semi-
structured interview are that the respondent is not limited by a set of predetermined answers,
like a structured interview. Furthermore, all interviews have been personal face- to- face
interviews. By performing face- to face interviews the researchers believe that non- response
can be minimized and questions can be explained further if the respondents have difficulties
grasping the meaning of a question. This will consequently increase the quality of the data
collected. Communication through telephone and e-mail has only been used when there has
been a need to clarify or supplement earlier interviews.
Methodology
53
In order to prepare for each interview the researchers prepared questionnaires (see appendix)
for each participants which was emailed to them one week before the interview took place.
This allowed the respondents to prepare for the topics of discussion. Moreover, all
interviews were tape-recorded. According to (Saunders, Lewis and Thornhill, 2003) the use
of a recorder allows the interviewer to fully concentrate on the respondent, become more
involved in the discussion and make observations. In addition, the researchers had the
opportunity to listen to the recordings afterwards to adjust any misunderstandings or
misinterpretations.
3.6.4 Secondary data Secondary data refers to data that already exists and is collected for another purpose. Most
research studies require secondary data for the background information. The quality of
secondary data depends on the source and method of presentation. By reviewing the quality
of evidence that has been presented in the arguments, the validity of the arguments and the
qualification of the presenter one can determine the quality of secondary data. This will
enable a researcher to identify bias and inaccuracies of the data. Secondary data consist of
literature, articles and reports. The researchers used secondary data as a complement to the
primary data. A wide extent of literature has been used to create the theoretical framework.
This includes books, scientific journals as well as articles and Internet sources such as
company web page. All internal material was provided by the company, including numerical
information regarding available containers during the last two years for ports located in
Helsingborg, Gothenburg, Norrkoping, Stockholm and Gavle, placed bookings during the
years of 2014 and 2013 for each above-mentioned port and import volumes for each port
during the same years. It is worth mentioning that the authors experienced issues with
gathering correct secondary data from the investigated company, this will be discussed more
in detail in the next section.
3.7 Research evaluation- reliability and validity Reliability and validity are vital concepts of a research as they measure the quality and
accuracy of the research findings (Druckman, 2011). Reliability is the extent to which other
researchers would gain the same findings if the study were repeated under the same
conditions (Karlsson, 2014). It is the consistency of measurement, the degree to which an
instrument measures the same way each time it is repeated under the same condition with
the same subjects. For a study to be considered reliable, a repeated study should produce the
Methodology
54
same results. Lack of reliability is due to inconsistencies in the research processes and
measurement errors. The level of reliability tends to be high in positivist studies, while
interpretive studies place little concern on the issue (Yin, 1994). It is argued that quantitative
research methods generate high reliability but lacks in validity. This is due to the fact that
focus lays on precision of measurement and the ability to repeat the experiment consistently
(Yin, 1994). Validity refers to how well a test measures what it is supposed to measure
(Bryman and Bell, 2003). How well the research results correctly reflect the subject of study
(Yin, 1994). When discussing validity one refers to internal validity and external validity.
Internal validity is the degree of accuracy in qualitative data and the extent to which one can
make a causal claim based on the study. External validity refers to how well the research
result can be generalized and applied to other people or settings (Bryman and Bell, 2003). A
researcher should start by establishing the internal validity in order to clarify which process
to explore and produce a conclusion. Followed by external validity, which seeks to define
the extent to which the results of the study can be generalized (Druckman, 2011).
By implementing a mixed methods approach, the researchers have incorporated methods
from both quantitative and qualitative research approaches in a single research study. The
data quality was determined by standards of quality in quantitative and qualitative research
instruments. The quantitative approach of this study allowed the researchers to use statistical
formulas that were unbiased and not possible to influence. Thus, enable other researchers to
reach the same findings, where the study to repeat in the future using the same statistical
formulas. Hence, facilitate high reliability.
The validity of this study, how well the research study measures what it is supposed to
measure, can be challenged. The researchers believe that the connection between the
theoretical framework and empirical data is strong. The qualitative approach enabled the
researchers to gain full knowledge of the subject, by extracting data that provided rich and
detail description regarding the subject at hand. However, when working with the statistical
data some errors appeared. Data regarding available containers were equal to zero during
some examined periods. This is however not consistent with reality. There have been no
periods during the last five years where available containers in Sweden have been equal to
zero. This proves that the data has been subject to error, thus decreasing the validity of the
study. The researchers have tried to extract correct data a second time in order to increase
the validity of the study. Nonetheless, this wasn’t possible and all calculations where hence
Methodology
55
based on existing numbers. The strength of the external validity, the extent to which the
study results can be generalized to other settings can also be discussed. By implementing a
mixed method strategy the researchers have been able to make use of both quantitative and
qualitative data. Still, the collected data is based on the case company and might not be
relevant to other companies or industries. Since all conclusions are based on the empirical
findings of the case company, and no studies have been made on other companies, the
researchers lack knowledge in regard to how well the results can be generalized to other
settings.
Empirical findings
4. Empirical findings This chapter presents the empirical findings. The objective of empirical findings is to capture and present the case study of the thesis. The author’s start by presenting company X followed by the results of the interviews conducted with different divisions of the organisation. It’s worth mentioning that company X don’t have a theoretically established forecasting process. How forecasting is done is internally created and adapted to the everyday operations. The layout of this chapter will follow the structure of the companies forecasting procedure for the convenience of the reader.
4.1 The case company The company X is a privately owned global shipping company. The company was founded
in 1970 and has grown from being a small tramp operator with one- second hand ship, to
becoming one of the world's leading container shipping companies. At current state
company X is operating a network of over 480 offices in 150 countries. They have an
established fleet consisting of approximately 500 container vessels with a capacity of 2 439
000 TEU. Their sailing schedule covers 200 routes calling on over 300 ports around the
world. In addition to offering ocean carriage the company has established a comprehensive
infrastructure of road, rail and warehousing. Enabling them to provide their customers with a
full transport solution (company webpage, 2015).
Company X is a country agency offering ocean services from the major Swedish ports:
Helsingborg, Gothenburg, Norrkoping, Stockholm and Gavle. In addition to ocean freight,
intermodal services are provided enabling cargo to be delivered door-to-door. The Swedish
agency consists of several departments, working hand in hand in order to create the best
possible transport solution for their cargo customers (Logistics interview). Each department
plays an important role in the creation of the final product, thus cooperation and information
distribution is of great importance. Henceforth, focus will be on explaining how different
departments conduct their everyday business in relation to the company forecasting process.
4.2 Information distribution design The information distribution design represents how the information flow is distributed
between the logistics department and other departments in regard to the companies
forecasting process. The model is illustrated as follows:
Empirical findings
57
Figure 10. Information distribution design at company X The role of the logistics department is central as they are responsible for conducting
forecasts related to container volumes. The logistics division is liable for providing all
departments with relevant information regarding container availability in each port. In turn,
they must access information from the other departments to make forecasts as accurate as
possible. The export division provides information regarding export bookings, more
specifically, container type and volume that are requested in specific areas. Further, the
import administration delivers information regarding import volumes that are to be returned
in each port. Finally, the logistics department is provided with relevant information
regarding different contracts and expected volumes from the commercial export department.
This can for example be specific tender agreement that will occur during a certain time
period.
4.3 Company X`s forecasting process The logistics department is responsible for the forecasting process at company X. At current
state company X does not have a theoretical established forecasting process. The forecasting
procedure is internally created and adapted to the everyday business operations.
Empirical findings
58
Consequently, this means that no theoretical forecasting technique is applied. The primary
objective when doing a forecast is to be as accurate as possible, which would result in
satisfying all bookings while keeping as low as possible container stock level in each port.
According to the logistics department the forecasting process is based merely on a short-
term time horizon meaning that predictions are done on weekly basis. Thus, enabling the
company to cope with the ever- changing shift in demand and supply. When predicting
container volumes the logistics department considers placed bookings one month in advance
and compare these figures to available containers in each depot/ port and expected import
volumes. The difference between the variables indicates whether there is a deficit or a
surplus in container volumes.
Deficit / surplus = available containers + expected import volumes - bookings
If a deficit of container volumes occurs, company X will not able to satisfy all bookings
resulting in a reduced customer service level. On the other hand, having a constantly
increasing surplus of containers in ports will increase container- holding costs (Logistics
interview).
4.3.1 Variable: Available containers Company X is very much directed by their head office in Geneva, they must always stay
within certain restrictions regarding safety stock levels. This means that when facing trade
offs between delivering high service level and keeping low stock levels, the aim is to deliver
the best possible service level while staying within the margins provided by Geneva. This is
a constant struggle for the logistics department as they are continuously trying to satisfy
demand from the different departments within their home office while meeting the
requirements provided by Geneva (Logistics interview). This can be illustrated through an
example:
During a four- week period the logistics department can note that customers have placed
bookings equivalent to 150 x 40 HC container in Gothenburg. Consequently the logistics
department makes an equipment request to their head office in Geneva asking for 180 x 40
HC, with 30 x 40 HC as a buffer for potential spot bookings. The head office notes that the
agency has 50 containers available in port and another 90 containers with import customers,
which they consider as soon to be available containers. Thus the head office will only
provide the agency with approximately 40 containers, since that is enough to stay within
Empirical findings
59
appropriate stock volumes. This however can create a deficit of containers in Gothenburg, as
it is unknown to the logistics department when the import volumes will return. Deepening
on who the import customer is and what kind of contract they have with company X the
import flows can vary greatly. If the import volumes are not returned within the needed time
frame the logistics department will experience a deficit of approximately 90 containers.
Thus, preventing them to satisfy all their bookings (Logistics interview).
The number of available containers is related to the actual number of containers in ports,
import volumes from previous periods as well as empty containers provided by the head
office in Geneva. Once a deficit of available containers occurs in Sweden, the head office
will regulate the situation by sending empty containers to the deficit areas (Logistics
interview). Consequently, once an area experiences a surplus of containers exceeding
existing bookings, these volumes are repositioned to areas around the world. Furthermore,
the cost of keeping safety stock varies in different ports. Company X has entered agreements
with the collaborative ports regarding number of containers that can be kept in the ports
storage free. Once this limit is exceeded the company will start evacuating containers to the
continent or other deficit areas within the country. An additional component affecting stock
level is seasonality. The shipping industry is very much affected by seasonality, which can
cause large fluctuations in demand for container volumes during a year. Accordingly this
will have an impact on stock levels (Logistics interview).
Company X experiences most export peaks before holiday seasons, such as periods before
Christmas and summer holidays. These peaks are predictable and the company can adapt in
advance to increase in demand. However, the biggest challenge when doing forecasts is spot
bookings, bookings that are placed with one or only a few days notice, and are
unpredictable. Number of spot bookings can vary greatly and become difficult to predict,
thereby difficult to cover (Logistics interview). At the same time its not possible for
company X to force their customers to place booking with better anticipation, this is because
company X goal is to be as flexible as possible to its customers. This means to resolve all
customer bookings although these are added in the last second (Commercial trade export
interview). How available container volumes have varied during 2013 and 2014 is illustrated
in following graphs:
Empirical findings
60
Graph 1. Available container volumes, Sweden 2013 & 2014.
Number of available containers is much dependent on import volumes and export bookings.
The number of available containers is positively related to increased import volumes, while
on the other hand the number of available containers decreases when export volumes
increase. Prediction of available container volumes is difficult as it is directly related to
these two highly oscillated variables. The graphs illustrate number of available containers
(20 DC, 40 DC and 40 HC) during 2013 and 2014. In the first graph one can see an increase
of volumes for 20 DC in 2014 during week 20. This is equivalent to increased import
volumes during the same period, as a result of increased shipping demand before the
Chinese New Year. In 2013 for the same period and container type the data has been subject
to error and therefore no valid result can be concluded. The next peak is during the weeks of
30-35. The increased number of available containers is due to increased number of import
container volumes during the months of April, May and June. Similar situation occurs
during the weeks 42-45 in 2014 and weeks 42-48 in 2013. Same seasonality variations occur
for 40 DC and 40 HC for the same periods of time.
4.3.2 Variable: Bookings Export bookings are an imperative variable when doing forecasts. Company X experience
difficulties in predicting the number of placed bookings in different periods, as several
factors determine export volumes. One of these factors is different type of customers (Pan,
2015). There are mainly two types of customers, customers entering tender agreements and
customers placing spot bookings. Tender customers, are customer who have entered an
agreement with company X to transport a large volume of cargo during a certain period of
time, however, the quantity per month is not defined in advanced. This information is
supplied on an irregular basis, normally a few weeks in advanced. Spot booking customers,
are customers who place a booking a day or a few days before vessel closing (Commercial
Empirical findings
61
trade export interview).
The split between bookings is very complex to predict. Still, the majority of bookings are
spot bookings. One can estimate that 60 per cent of all bookings consist of spot bookings.
This is due to different reasons, including macro-economic reasons that neither the company
nor the customer can predict. In addition, tender customers also have an ability to make
“spot bookings” (Export interview).
Tender customers most often inform how many containers that will be transported in a year,
however they don’t give details on how the distribution is divided over a year. It’s very
common for a tender customer to “lock” half of their volumes to pre determine dates, while
the other half is often placed with a few days of notice (Commercial trade export interview).
Still, when looking at historical data one can see that peaks and lows for container volume
are somewhat similar. This can be illustrated by the graphs:
Graph 2. Export bookings, Sweden 2013 & 2014 A compilation of booking volumes for Gothenburg, Helsingborg, Norrkoping, Stockholm
and Gavle is shown in the graphs for 2013 and in 2014. It appears that the trend for both
years is very similar. It can be seen that most peaks in both years are during the months of
March-April, June, September and December for both 40 HC and 20 DC containers in
Sweden. 40 DC on the other hand has a clear peak during the months of January, July and
September. The reason for this is that most peaks coincide with various holidays. For
example export during June is very high just before many industries takes a summer break.
Another peak is in September- October, as many companies want to export their products in
time for the Christmas holidays.
All information regarding bookings is provided to the logistics department as soon as it
Empirical findings
62
becomes available. However, the information is subject to time lags. When a customer
places a booking it has to be processed by the export department. This means that it may
take a day or longer before information becomes available to the logistics department
(Export interview). The goal is to always provide information as soon as possible; this can
however be complex as it depends very much on the customer. The export department does
not work proactively towards customers, rather, they work reactive which means that
customers reach out to company X with everything booking related. Even if the export
department worked proactively to get booking information earlier, the fact of the matter is
that the customer in most cases does not have the information to give as they haven't
received the information in turn from their customer (Commercial trade export interview).
4.3.3 Variable: Import volumes Import volumes are an essential variable for predicting container volumes. To assess how
many containers will be available in order to satisfy customer bookings, a prediction of
import volumes is necessary. There are different types of customers that order shipment
services from company X and the balance of import volumes differs for different markets
(Import interview).
Customer prices vary during a year, as some costs are fixed or relatively stable, while others
fluctuate over time. Freight prices for the sea transportation vary approximately 42 times per
year. Shipping demand is highly correlated to the freight price level, however this variable is
not used when estimating import volumes. The continuous fluctuation of freight prices is
based mainly on the competition between different shipping companies as well as on the
fluctuation in fuel rates. Shipping companies can lower freight rates in order to attract more
customers. This pricing competition can continue until the rates are to low to generate any
profit. This will consequently result in a slow increase in price in order to get the market
back on track (Import interview).
There are peaks in demand for import shipping, which is mostly related to the Far East trade
as 45 per cent of the all import container volumes come from this area. This can be
illustrated through following graphs of imported volumes for 2013 and 2014 in Sweden:
Empirical findings
63
Graph 3. Import volumes, Sweden 2013 & 2014
The biggest peaks in import volumes are in April and May as a result of the Chines New
Year with an additional peak in September and October when companies start to order for
the Christmas period. In 2014, 20 DC containers experience most peaks during April, May
and June and September. This is consistent with the import volumes in 2013, while 40 HC
have a very strong peak during the month of June. In 2013 however, it appears that import
for 40 HC have been relatively stable with one peak in September and then slowly subside.
Volumes for 40 DC is consistently low both for 2013 and 2014.
Due to the fact that company X agencies around the world don't use the same operative
systems, the agency in Sweden do not receive any information regarding number of import
volumes in advance. Once the containers are loaded on board the ship, approximately one
week before reaching the Swedish ports the information becomes accessible to the import
department (Import interview). When containers are in land in Sweden, a customer can keep
a container as long as needed, however after seven days most customers will be charged
with demurrage free for each exceeding day. Some customers have special agreements with
company X, which means that they can keep a container during a longer time period. This
has however made prediction of import volumes very difficult.
Analysis
5. Analysis
This section of the thesis includes an analysis of the literature review and the empirical findings presented in previous chapters. As the purpose of this paper is to “evaluate the existing forecasting process and accuracy of forecasting techniques implemented by company X, in order to identify and propose an improved forecasting process for predicting container volumes”, an analysis of collected data using theories discussed in the literature review will be performed.
5.1 The importance of a forecasting process The literatures propose that a forecasting process is an essential activity in many business
organizations. Not only does the activity allow companies to predict future conditions and
eliminate potential risks, but also because it supports businesses to reach sustainable growth
opportunities (Jacobs, Chase and Lummus, 2011). The process has become a necessity for
organizations to cope with ever changing shifts in demand and supply (Thomopoulos, 1980).
The implementation of a tool that can generate up to date information is vital for every
significant management decision, as it eases business planning and makes it more efficient
(Diaz, Talley and Tulpule, 2011). The literature review presents several important steps in a
forecasting process that needs to be identified in order to implement a well- constructed
process that can help companies improve their performance and competitive position in a
business market. The aim is to put these steps in relation to company X in order to identify
strengths and shortcomings of the companies forecasting activities.
5.2 The forecasting process of company X
5.2.1 Identify strategic goals The first step of a forecasting process should be the identification of strategic goals.
According to Makridakis and Wheelwright (1987) a good forecasting process is constructed
in alignment with the strategic goals of a company. By aligning the forecast process to the
strategic goals of the company it can be ensured that employees has adequate understanding
for why and how business should be conducted (Mintzberg, 1976). The incorporation of a
forecast activities in the organizational decision making process will allow effective trade
offs between operational activities, thus increase effectiveness and competitive advantage
(Schultz, 1992). The strategic goals of company X are to offer global service to all its
Analysis
65
customers while maintaining low costs. However the forecasting activities of the company is
not aligned with their strategic goals. This is due to the fact that there is no theoretically
established forecasting process at company X. How forecasting is conducted is internally
created and adapted to the business operations (Logistics interview).
Furthermore, according to Janeš and Faganel (2013) organizations use KPIs to monitor the
performance and realization of the organization`s strategy. The use of KPI is a way to
measure how well a company has succeeded in their performance in comparison to their
targets. All evidence in the empirical findings shows that company X has very clear KPIs
that are known to their employees. The company works continuously to successfully achieve
these (Logistics interview). Still, since the forecasting process is not aligned to the
company’s strategic goals and the strategic goals do not form the bases of the forecasting
process, it’s believed that the company lacks in implementing a well- constructed forecast. If
company X develop and align their forecasting activities to the strategic goals and
competitive ideas of the organisation, with participation of the people who really understand
the company`s competitive strategy, they will be able to implement a process that has clear
directives. The probability to successfully make predictions that coincides with the
company’s objectives will thus increase.
5.2.2 Decide the objectives of the forecast One important step in a forecasting process is the identification of the forecasting objectives.
According to Daim and Hernandez (2008) by defining the problem one will create an
understanding of the intended use of the forecast, thereby creating a better fit of the forecast
purpose to the organization requiring the prognoses. Once the objectives of the forecast are
clarified one will have a clear understanding of the level of details required. The forecast
must align with the target objectives in order to be accepted. According to the logistics
department at company X the primary objective of their forecasts is to conduct forecasts
with as high level of accuracy as possible. This means that prognoses on required container
volumes in each Swedish port should be as close as possible to actual bookings placed in
each port. Consequently, with a high level of accuracy the company will be able to eliminate
all costs related to potential surplus or deficit volumes. In the course of our research we
could see that the logistics department is very much directed by their head office in Geneva.
Geneva provides very clear limits regarding safety stock volumes allowed in each port. The
limits are extremely strict, and the logistics department must always stay within these
Analysis
66
restrictions. Hence, while company X has clear objectives for their forecast, due to
restrictions they are not always able to act accordingly. Thus making it very difficult to
achieve the forecast objectivise.
5.2.3 Choose a forecast approach In order to choose a preferable forecast approach, it is required to investigate if all necessary
conditions are fulfilled. When exploring historical data, statistical methods are used for
analysis of time series (Axsäter, 2006). Thus, a reliable database has to be accessible. At
current state, company X partly save information regarding previous demand. Historical data
regarding inventory levels, customers’ bookings and volumes of import containers are
available, but this set of measurements should be complemented with an additional form.
Information regarding cancelled bookings is not collected. This data would help a forecaster
to better understand special characteristics associated with the sequences of the
observations. When all data is gathered and completed, company X will more easily be able
to explore historical data and apply the results to the operational control systems.
Choosing one forecasting approach does not necessarily mean that the other one should be
excluded. As company X operates on the global market, demand for shipping services is
correlated to freight price. Shipping companies can decide to lower their freight rates in
order to attract more customers or increase freight rates when they see profit opportunities
emerge. By being able to predict freight price level on the global market, company X can
use this approach in order to forecast the future customers` demands. Once again, data
collection is a crucial step when deciding a forecast approach.
5.2.4 Choose forecast variable from collected data According to the literature the quality of a forecast is only as good as the quality of the input
data. A proper selection of variables is therefore necessary in order to conduct a forecast
process with high performance (Malakooti, n.d.). When observing company X forecasting
practise it was revealed that the company predicts three variables, available container,
existing bookings and import volumes. The company believes these to be an optimal set of
variables to capture the true relationship of the phenomena. Yet, by analysing company X
way of predicting container volumes, the authors have identified a fourth and a fifth variable
that needs to be considered, that is spot bookings and freight rates. Spot bookings have
become a constant struggle to predict and cause the forecast to decrease in accuracy. By
including spot bookings in the calculation one can generate a forecast with higher level of
Analysis
67
precision. The authors have also detected from the literature that freight rates has an impact
on demand for shipping services (Axsäter, 2006), thus this variable should be considered
when computing forecasts. One needs to keep in mind that too many variables may cause
the model to become over fitting, still the authors believe that the fourth and fifth variables
will have a strong correlation to the data set and therefore generate a forecast with high
accuracy.
5.2.5 Choose forecast horizon According to Cook (2006) forecasts are made with reference to a specific time horizon. The
time horizon of a forecast is the estimated length of time that a company decides to predict.
The choice of time horizon will consequently have an impact on the choice of forecasting
methods. According to the empirical findings the authors can deduce that the market where
company X operates is volatile with demand changing continuously. Due to high level of
uncertainty regarding future bookings, company X conduct their forecast on short- term
basis. Forecasts are made on weekly basis, also known as a tactical forecast for predicting
future container volumes.
The empirical study reveals that how far a plan extent into the future is dictated by the
degree of uncertainty. A great degree of uncertainty will require a longer planning horizon
and vice versa. By conducting forecasts on weekly bases the logistics department intends to
ensure sufficient lead- time to receive the requested equipment volumes. In addition, short-
term forecasts have proven to be more accurate than long-term forecasts as they have an
ability to adjust the existing plan based on new information obtained (Chan, Cheng and
Fung, 2010). Using a long- time horizon would only be beneficial for future predictions of
container volumes if these forecasts were used as a complementary to conducting short-term
predictions. After researching existing literature the authors have come to the conclusion
that for a given horizon of n periods, company X should allocate a schedule of container
volumes to satisfy the given demand in each period, thus keeping inventory cost as low as
possible. It is suggested that a time horizon with a relative stable interval should be selected,
thereby allowing the forecaster to hedge against fluctuations in the market.
5.2.6 Identify demand patterns and decomposition of data The theoretical framework presents several demand patterns that can be identified to ease
forecasting processes. When performing forecasts it is useful to make assumptions regarding
how the relationship between different variables affecting demand will continue in the
Analysis
68
future. Newbold, Carlson and Thorne (2013) have identified and presented four main
patterns of time series data that forecasting is based on. These demand patterns can
subsequently be identified and separated by using decomposition or simply by plotting the
data. There is no empirical evidence that company X implement any theoretically explained
models. Identification of demand patterns is done in some small extent, however
decomposition of the data is not implemented at all. When doing forecasts some
considerations are taken to historical trends, still this action is not taken continuously each
time a forecast is made. The authors believe this to be a disadvantage for company X. If the
company would keep track of how their demand patterns evolve over time, they will have
the ability to adjust quickly to changes in demand. This will have a direct impact on the
company's stock level, inventory cost and subsequently their service level.
Company X should analyse the data in order to get some understanding of the data`s
possible variations through time. The authors recommend company X to do decomposition
of the data using software R, as the procedure is simple, the software is free and it is
available for every operative system. This procedure was used in the later stages of the
research for calculating smoothing parameters alpha and beta, when Holt-Winters non-
seasonal method was computed and tested.
5.2.7 Choose forecasting technique and error measurement According to Jarrett (1987) there are several existing forecasting techniques that can be used
to predict future numeric estimates. These techniques range from relatively simple to
complex methods. The selection of a forecasting technique is dependent upon availability of
historical data, degree of accuracy desirable and available time to perform the forecast. The
technique that is believed to provide the greatest benefits for the company should
consequently be chosen (Chambers, Mullick and Smith, 1971). Consistent with the logistics
department of company X there is no theoretically compatible forecasting technique
implemented when doing future predictions. The way forecasting is done is internally
created (Logistics interview). Even though company X do not employ any of the techniques
mentioned in the literature review, the authors of this study believe that a comparison of
calculations based on the data provided by company X using the techniques from the
literature, can be useful. Thus, a calculation of company X data using the techniques moving
average, Holt- Winters non- seasonal method and simple regression analysis will be
conducted and presented. The techniques have been randomly selected; hence no underlying
Analysis
69
intention lies behind the choice. Moreover, the accuracy of the named forecasting techniques
will be tested and compared in order to gain some basic understanding for the performance
of the forecasting techniques.
Period 40 HC
demand
Forecast 40 HC demand using
moving average
Forecast 40 HC demand using Holt-
Winters non seasonal method
January 1696
February 1874
March 2142
April 1763 1904 2107
May 1767 1926 2011
June 2107 1891 1863
July 1731 1879 1979
August 1638 1868 1828
September 1900 1825 1629
October 1708 1756 1695
November 1491 1749 1672
December 2100 1700 1514
𝛂 0.3612355
𝜷 1
Table 2. Forecast of 40 HC demand using moving average and exponential smoothing method
MAD MSE MAPE
Moving average 186 50199 10.3 %
Holt-Winters non seasonal
exp. smoothing
258 98153 13.99 %
Table 3. Evaluation of the forecasting results
Analysis
70
The numerical evidence for the moving average technique and Holt-Winters non seasonal
exponential smoothing method have been calculated manually, while the smoothing
parameters alpha and beta have been generated via software R by using HoltWinters (x,
alpha, beta, gamma) function. By computing the empirical data for 40 HC containers using
the error measurement formulas, one is able to compare the accuracy of the alternative
forecasting techniques. It appears from the calculations that the moving average technique
yields lower results for MAD, MSE and MAPE, compared to Holt-Winters non- seasonal
exponential smoothing method. According to the generated results the authors can conclude
that moving average is the best suited technique for predicting future container volumes
between the two techniques, as moving average generates the most accurate results.
Simple regression analysis In order to compute a simple regression analysis, the author’s explored parameters that
could affect the demand for export bookings as it might be related to different variables.
According to Axsätter (2006), one of the most common variables that affect demand of a
product or service is price level. The demand for a shipping service goes up when the freight
rates go down and reverse. In absence of freight rate data, the authors identified another
variable that might indirectly affect the demand for export bookings, as it is directly related
to the freight prices. When company X has a surplus of available container volumes, they
may decrease their freight rates in order to decrease the container volumes in ports. Thus, an
increase in available container volumes might lead to increased demand for export bookings.
Before calculating the correlation between these two variables, the authors concluded two
hypotheses:
Ho = There is no relationship between the two measured variables available containers and
export bookings.
H1 = There is a relationship between the two measured variables available containers and
export bookings.
These hypotheses are tested and the correlation level is presented in the following graph:
Analysis
71
Graph 4. Simple linear regression technique for 40 HC, Sweden 2014
It can be read from the plot that the correlation between export bookings and available
container volumes is quite low (R2= 0,4149) proving that the null hypothesis (H0) is correct.
Surprisingly, the slope of the simple linear regression line is negative (b = - 0,2138), which
means that when available container volumes increase, the export bookings will decrease.
This finding provides us with an additional evidence of how the shipping market is difficult
to predict. In conclusion, the authors noted that it would have been favourable to test the
correlation between prices and export bookings, as these two variables would probably show
the strongest correlation.
5.2.8 Evaluation of forecasting techniques & results Evaluation of forecasting results is not a simple task, as it doesn’t always mean choosing the
most accurate forecasting technique, while ignoring other factors (Chatfield, 2001). Besides
the importance of choosing the most accurate forecasting method, the judgmental abilities
play a crucial role in the process of evaluation. Once the different forecasting techniques had
been tested, the authors were able to calculate the most common statistical error
measurements such as MAD, MSE and MAPE, which were presented in the previous
chapter. Once accuracy of the forecast has been calculated, company X should proceed with
the judgmental process. This activity should involve the person computing the forecasts and
the department using the forecast. The main focus of the evaluation process at company X
should be on the acquisition of the forecasting results, how these results are processed and
output on the individual level. Moreover, company X should take into consideration actions
taken as a result of the forecaster`s beliefs and the final outcome of these actions. Increased
sales, costs and ultimately profit are just some of the outcomes that should be analysed. In
conclusion, a proper feedback should reach the logistic department at company X, in order
Analysis
72
to improve and monitor the existing forecast process.
5.2.9 Integration of forecast process According to the literature, forecast integration is an essential step of the forecast process. A
successful integration of a forecast process across an organization can significantly improve
a company's performance, as it increases transparency, information sharing and provides
support to making decisions (Pal Singh Toor and Dhir, 2011). Some scientists have
discussed the complexity of integrating a forecast process, as it requires high managerial
skills. Still, emphasis is put on the fact that the high level of complexity enhances
importance of the integration process, as good forecasts can sometimes fail through poor
planning and their implementation (Makridakis and Wheelwright, 1987). All empirical
evidence points to the fact that company X do not integrate their forecasts or their forecast
results across their organisation. In fact, it was proven that company X must make great
improvements in this area. The forecasting process is merely integrated within the logistics
department, as only the logistics department carries out the forecasting actions. This has
accordingly created a gap within the company where different departments lack knowledge
and information regarding the forecasting actions completed.
The authors did also detect lack of information sharing between the departments at company
X as well as, between company X, their head office in Geneva and other agencies around the
world. This has diminished transparency across the organization, as some important
information does not reach company X as quickly as it could. If company X would commit
to integrating their forecast process across their enterprise, so all departments have the right
knowledge and understanding of what a forecast process involves, the company would
enhance several aspects of their operations. Increased information sharing and transparency
would help the company to make decisions that would generate the most beneficial results.
They would be able to improve recording capabilities for data, where the company could
analyse their actions in relation to their KPIs. By integrating the forecast process in all
aspects of the company company X can work towards increasing the efficiency of their
operations in order to achieve higher profitability. In addition, monitoring forecasting results
will allow company X to have a sustainable forecasting process that will yield indefinite
improvements. Thus, there is a necessity for an on-going forecasting process.
Analysis
73
5.3 Comparison between company X `s and the researchers proposed forecasting process Below a comparison between the researcher’s proposed forecasting process and company X
´s forecasting activates are presented. Company X agrees that a well-developed forecasting
process is a necessity for every day operations, but still no theoretically established
forecasting process is applied. Table 4 presents all steps identified by the researchers as a
necessity for conducting a sustainable forecasting process, in relation to actual measures
taken by the case company.
The researchers proposed forecasting process
The internally created forecasting process of company X
Identify strategic goals
Decide objectives of the forecasts
✓
Choose a forecast approach
Choose forecast variables from collected data
✓
Choose forecast horizon
✓
Identify demand pattern
Choose forecast technique
Evaluation of forecasting techniques and results
Integration of forecast process
Table 4. A comparison between company X`s and the researchers proposed forecasting process
Conclusion
6. Conclusion
In this final chapter the authors will draw conclusions that are based on the result of the analysis. The aim is to answer the research question “How should a forecasting process for predicting container volumes at company X be designed, in order to generate accurate forecasts?” and provide suggestions for improvements.
The shipping industry is a highly volatile and uncertain industry and if not predicted
accurately, can cause financial instability for organizations. A proper-implemented
sustainable forecasting process is therefor essential in order to adapt to continuously
changing trends and strengthen operational management. The implementation of a well-
structured forecasting process is fundamental for every significant management decision, as
it provides its executives and management with a proper tool to improve their performance
and competitive position. The authors of this paper have created and presented a sustainable
forecasting process that goes beyond using and developing accurate techniques, filling the
gap provided by the existing literature. The proposed forecasting process allows forecaster
to conduct forecasts by aligning it to the strategic goals, but also to evaluate, monitor,
integrate and improve forecasts through time. These features added to the fact that no step in
the process is underestimated, may allow forecaster to conduct more accurate and
sustainable forecasts. The proposed forecasting process was challenged in this thesis by
testing it in the highly uncertain shipping industry, something that has once again been
avoided by the existing literature. The analysis of the investigated subject revealed that
adapting all steps of the proposed forecasting process would ease future predictions. The
first conclusion that can be made thereof is that company X should apply all the steps of the
presented process for future forecasting activates, as each step builds on the previous one
and can together create sustainable results.
It has become clear to the researchers of this study that the maritime industry is a highly
problematic industry to predict, characterized by quick changes in supply and demand. The
ability to anticipate market movements can however increase with the use of a forecasting
process that is aligned to the strategic goals of the company. Hence, the first step for
company X is to align its forecasting process to the overall strategic objectives of the
Conclusion
75
corporation, so that expected trade offs between customer service level, inventory level and
procurement economics can easily be interpreted. If there is lack of alignment, company X
can spend large amount of time reacting to crisis, instead of preparing for these sudden
predicaments. In addition, company X needs to identify which activities that add value to
their forecasting process, and align these to the company strategy. By doing so company X
will acquire clear directives for conducting their forecasting activities, thereby experiencing
enhanced inventory management and reduce tied up capital in stock level.
An additional key to conducting a successful forecasting process is the element of
information gathering and information sharing between departments. It has become evident
that important information regarding forecast activities is available in various departments
within company X, both internally and externally, however this information is not shared or
utilized in an effective way. In order to increase visibility of information it is recommended
that company X integrate a synchronized IT system that is available for all company X
agencies around the world. This can be done for example by using an intranet based system
where customer themselves can manually insert bookings. By doing so the logistics
department within company X will be able to access information regarding placed bookings
and import volumes back to Sweden sooner than currently received. The increased visibility
will subsequently allow company X to improve their planning of container volumes, rather
then forcing them to react to sudden container shortage. If company X follows the proposed
suggestions they will reduce cost in tied up capital and experience increased service level.
Increased collaboration between company X agencies around the world is a necessity, as the
competition in the future will depend on the performance of the whole supply chain, and not
the individual performance of each department. It is worth mentioning that all improvements
for forecasting will lose its effects without co- operation between departments within and
outside of Sweden, it is a crucial success factor.
Additional ways for increasing accuracy is to include customer relationship management as
an additional step in the forecasting process. By enhanced collaboration with customers,
company X can collect appropriate and valuable data in advance that will increase
trustworthiness and reliability of the forecasting process. Indicators for ordering quantities
and timing variations are needed to ease the forecasting process and remove uncertainties.
At current state company X is conducting a very reactive customer relationship
management, which means that information becomes available once the customer reach out
Conclusion
76
to the company. If company X would work proactively they would be able to increase
visibility of their customer’s wants and needs. Thus, minimizing problematic spot bookings
and prediction of point of container return. Moreover, an enhanced customer relationship
management would provide continuous information flow, thus enabling company X to
gather necessary data in advance. This will additionally enable company X to conduct long-
term forecast that is more consistent with reality and can be used as a complement to the
already existing short-term forecast.
Once all essential data is collected it needs to be properly processed. This can be done with
the use of various softwares. Currently company X lacks software were predictions of
container volumes can be done. A suggestion is therefore for the company to acquire a
computer software program that can translate data into future predictions, such as software
R. With the use of such software, company X can easily decompose data in order to identify
demand patterns for containers. Once predictions have been made and proper error
measurements have been conducted, company X should proceed with the judgmental
process of evaluating and monitoring the forecast predictions, this includes sales, costs and
profit.
The forecasting process should be viewed as an on-going process as each step builds on the
previous one and together create sustainable results. This means that once the final step of
the process is reached it is of great importance that feedback regarding outcomes reaches the
logistics department where the forecast is conducted. By monitoring forecasts and providing
the logistics department with relevant feedback regarding outcomes, the forecaster will
recognize if the conducted forecast is sufficient or if adjustment needs to be made in order to
improve forecasts. Hence, it is crucial that the forecasting process is integrated across the
entire organization and the final results of the forecast reach all departments. Thus,
providing all departments with a proper knowledge and understanding for what a forecast
process represents. By integrating a forecast process across an entire enterprise, company X
will be able to ease evaluation process based on true economic impact, improve strategic
alignment and position, increase transparency, provide decision support, optimize asset
utilization and improve financial performance. Thus, conducting an on-going forecasting
process that will result in continuous improvements. In conclusion following sustainable
forecasting process is recommended:
Conclusion
77
6.1 Future research suggestions This research thesis has highlighted the concept of forecasting as a process, where main
focus is on presenting and investigating all necessary steps required conducting an efficient
forecast. The research study has identified a number of areas where a future analysis would
be constructive. In order to satisfy the request of the case company and narrow down the
scope of the research the analysis was based only the shipping industry. One suggestion for
future research is therefor to apply the proposed forecasting process to other industries to see
whether it is applicable or not. It is also suggested to investigate whether the forecasting
process is applicable to other shipping companies within the industry.
In addition, focus has mainly been on quantitative and casual forecasting techniques,
excluding qualitative forecasting methods from the research scope. An additional research
Identfy strategic goals
Decide the objectives of the
forecast
Choose a forecast approach
Choose forecast variables from collected data
Choose forecast horizon
Identify demand pattern
Choose forecast technique
Evaluation of forecasting
technique & results
Integration of forecast process &
results
Conclusion
78
suggestion is therefor to examine if qualitative forecasting methods are appropriate for
future predictions of container volumes in the shipping industry.
Furthermore, it would be of great significance to study different pricing strategies and the
impact of pricing as a variable for future predictions. Due to confidentiality agreements and
time restrictions it has not been possible for the researchers to investigate the influence of
price on container demand. However, it is believed that price has a significant impact on
demand and should thereof be analysed. Thus, there is an identified need for more research
in the field of forecasting.
References
7. References Ambrosius, W. (2007). Topics in biostatistics. Totowa, N.J.: Humana Press. Axsäter, S.
(2006). Inventory control. New York: Springer.
Babin, B. and Svensson, G. (2012). Structural equation modeling in social science
research.European Business Review, 24(4), pp.320-330.
Barry, M., Devitt, J. and Ohno, R. (2015). Logistics Department. Gothenburg, Sweden
Beenstock, M. and Vergottis, A. (1993). Econometric modelling of world shipping. London:
Chapman & Hall. B illah, B ., K
smoothing model selection
for forecasting. International Journal of Forecasting, 22(2), pp.239-247. B rockw ell, P
Davis, R. (2010). Introduction to time series and forecasting. New York [u.a.]:
Springer. B rooks, C . (2002). Introductory econometrics for finance. Cambridge:
Cambridge
Bryman, A. (2007). Barriers to Integrating Quantitative and Qualitative Research. Journal of
Mixed Methods Research, 1(1), pp.8-22.
BusinessDictionary.com, (2015). What is forecasting? definition and meaning. [online]
Available at: http://www.businessdictionary.com/definition/forecasting.html [Accessed 5
May 2015].
Carson, D. (2001). Qualitative marketing research. London: SAGE.
Chan, W., Cheng, X. and Fung, J. (2010). Forecasting volatility: Roles of sampling
frequency and forecasting horizon. J. Fut. Mark., 30(12), pp.1167-1191.
Chambers, J., Mullick, S. and Smith, D. (1971). How to choose the right forecasting
technique. Harvard Business Review., Vol. 49(Issue 4), pp.p45-70. 26p
Chatfield, C. (2001). Time-series forecasting. Boca Raton: Chapman & Hall/CRC. C ho, J.
(2006). Validity in qualitative research revisited. Qualitative Research, 6(3), pp.319-340.
References
80
Collis, J. and Hussey, R. (2003). Business research. Houndmills, Basingstoke, Hampshire:
Palgrave Macmillan.
Cook, A. (2006). Forecasting for the pharmaceutical industry. Aldershot, England:
Cortinhas, C. and Black, K. (2012). Statistics for business and economics. Hoboken, N.J.:
Wiley.
Dagum E. B. (2010), TIME SERIES MODELING AND DECOMPOSITION,
STATISTICA, n. 4, 2010
DAIM, T. and HERNANDEZ, A. (2008). A FRAMEWORK FOR MANAGING THE
FORECASTING PROCESS. International Journal of Innovation Management, 12(04),
pp.597- 627.
Denzin, N. and Lincoln, Y. (2005). The SAGE handbook of qualitative research. Thousand
Oaks: Sage Publications.
Diaz, R., Talley, W. and Tulpule, M. (2011). Forecasting Empty Container Volumes. The
Asian Journal of Shipping and Logistics, 27(2), pp.217-236.
Druckman, J. (2011). Cambridge Handbook of Experimental Political Science. Cambridge:
Cambridge University Press.
Ferbar, L., Čreslovnik, D., Mojškerc, B. and Rajgelj, M. (2009). Demand forecasting
methods in a supply chain: Smoothing and denoising. International Journal of Production
Economics, 118(1), pp.49-54.
Gardner, J., Rachlin, R. and Sweeny, A. (1986). Handbook of strategic planning. New York:
J. Wiley.
Granger, C. and Pesaran, M. (2000). Economic and statistical measures of forecast
accuracy.Journal of Forecasting, 19(7), pp.537-560.
Gratton, C. and Jones, I. (2004). Research methods for sport studies. London: Routledge.
Greene, J. (2007). Mixed methods in social inquiry. San Francisco, CA: Jossey-Bass.
Harrell, F. (2001). Regression modeling strategies. New York: Springer. G uba, E. G .
References
81
Lincoln, Y. S. (1994). Competing paradigms in qualitative research. In: N. K.
Denzin H ow e, K . (1992). G etting over the Q uant-Qualitative Debate. American
Journal of
Education, 100(2), p.236. H yndm an, R . and A than Forecasting.
Heathmont, Vic.: OTexts.
Hyndman R. J. and Athanasopoulos G., “ Forecasting: principles and practice”, downloaded:
15- 04-07, https://www.otexts.org/fpp/6/4
Jacobs, F., Chase, R. and Lummus, R. (2011). Operations and supply chain management.
New York: McGraw-Hill Irwin.
Janeš, A. and Faganel, A. (2013). Instruments and methods for the integration of company's
strategic goals and key performance indicators. Kybernetes, 42(6), pp.928-942.
Jarrett, J. (1987). Business forecasting methods. Oxford, UK: Blackwell. Karlsson, B.
(2015). Research Methods Lectur Theory and Practice.
Kauko, K. and Palmroos, P. (2014). The Delphi method in forecasting financial markets—
An experimental study. International Journal of Forecasting, 30(2), pp.313-327.
Lin, V. and Song, H. (2014). A review of Delphi forecasting research in tourism. Current
Issues in Tourism, pp.1-33.
Lee Rodgers, J., Beasley, W. and Schuelke, M. (2014). Graphical Data Analysis on the
Circle: Wrap-Around Time Series Plots for (Interrupted) Time Series Designs. Multivariate
Behavioral Research, 49(6), pp.571-580.
Magaji, F. (2015). Commercial Export Department. Gothenburg, Sweden Maggi, G. and
Skoglund, M. (2015). Import Department. Gothenburg, Sweden
Makridakis, S. and WHEELWRIGHT, S. (1987). The handbook of forecasting. New York:
J. Wiley.
Malakooti, B. (n.d.). Operations and production systems with multiple objectives.
McKenzie, E. and Gardner, E. (2010). Damped trend exponential smoothing: A modelling
References
82
viewpoint. International Journal of Forecasting, 26(4), pp.661-665.
McKinnon, A. (2010). Green logistics. London: Kogan Page. M iddel R .,
University of Gothenburg, Gothenburg, Sweden, 2013-11-26, (lecture in Operational
Management).
Mintzberg H., (1976), Planning on the left side and managing on the right, Harvard Business
Review.
Montgomery, D., Peck, E. and Vining, G. (2012). Introduction to linear regression analysis.
Hoboken, NJ: Wiley.
Newbold P., Carlson William L. and Thorne Betty M., (2013), Statistics for Business and
Economics, Pearson Education Limited.
Ord, J. and Fildes, R. (2013). Principles of business forecasting. Mason, OH: South-Western
Cengage Learning.
Ostertagová, E. and Ostertag, O. (2012). Forecasting using simple exponential smoothing
method. Acta Electrotechnica et Informatica, 12(3).
Pan, D. (2015). Export Department. Gothenburg, Sweden Pal Singh To
(2011). Benefits of integrated business planning, forecasting, and
process management. Business Strategy Series, 12(6), pp.275-288. Piquero, A .
Weisburd, D. (2010). Handbook of quantitative criminology. New York:
Springer.
Reiner J., professor at Goethe University Frankfurt, Frankfurt, Germany, 2014-10-14
(lecture in Applied Marketing Research Methods)
Rowe, G. and Wright, G. (1999). The Delphi technique as a forecasting tool: issues and
analysis. International Journal of Forecasting, 15(4), pp.353-375.
Saunders, M., Lewis, P. and Thornhill, A. (2003). Research methods for business students.
Harlow, England: Prentice Hall.
Schultz, R. (1992). Fundamental aspects of forecasting in organizations. International
References
83
Journal of Forecasting, 7(4), pp.409-411.
Shahin, A. and Mahbod, M. (2007). Prioritization of key performance indicators. Int J
Productivity & Perf Mgmt, 56(3), pp.226-240.
Shalabh, (2013). A revisit to efficient forecasting in linear regression models. Journal of
Multivariate Analysis, 114, pp.161-170.
Shim, J. and Siegel, J. (1988). Handbook of financial analysis, forecasting & modeling.
Englewood Cliffs, N.J.: Prentice-Hall.
Shiskin, J. (1961). Tests and revisions of Bureau of the Census methods of seasonal
adjustments. [Washington]: U.S. Dept. of Commerce, Bureau of the Census.
Snyder, R. (2006), Discussion, International Journal of Forecasting, 22, pp. 673-676.
Stevenson, W. (2005). Operations management. Boston: McGraw-Hill. Stopford, M
(2009). Maritime economics. London: Routledge.
Taylor, J. (2008). Exponentially weighted information criteria for selecting among
forecasting models. International Journal of Forecasting, 24(3), pp.513-524.
Teddlie, C. and Tashakkori, A. (2009). Foundations of mixed methods research. Los
Angeles: SAGE.
Thomopoulos, N. (1980). Applied forecasting methods. Englewood Cliffs, N.J.: Prentice-
Hall.
Thomopoulos, N. (2015). Demand forecasting for inventory control.
Tran, H., Muttil, N. and Perera, B. (2015). Selection of significant input variables for time
series forecasting. Environmental Modelling & Software, 64, pp.156-163.
Venables, W. and Smith, D. (2009). An Introduction to R. United Kingdom: Network
Theory Limited.
Vlahogianni, E., Golias, J. and Karlaftis, M. (2004). Short‐term traffic forecasting:
Overview of objectives and methods. Transport Reviews, 24(5), pp.533-557.
References
84
Walliman, N. (2011). Research methods. London: Routledge.
Winklhofer, H., Diamantopoulos, A. and Witt, S. (1996). Forecasting practice: A review of
the empirical literature and an agenda for future research. International Journal of
Forecasting, 12(2), pp.193-221.
Yin, R. (1994). Case study research. Thousand Oaks: Sage Publications. Zucchini W. and
Nenadic O. ( n.d.), Time Series Analysis with R- Part 1.
Appendix
Appendix Interview questions for the Logistics department at company X
Forecasting process:
● Explain how the forecasting process looks today
● Do you believe your process is sufficiently clear? Can it be improved?
● What steps does it consist of?
Strategic Goals:
● What are the strategic goals of company X?
● Is the strategy of company X consistent with the forecasting process of the logistics
department?
● Companies can integrate their forecasting process with their strategy in order to
better meet company goals, it is something company X do?
● Do you use forecasting results as a basis for decision-making?
● Do you take into account trade offs between customer service level, inventory levels
(holding costs) & Cost efficiency? How so?
● Do you believe that your forecasting process can improve the efficiency of the
company and strengthen your competitive advantage?
● Do you use Key Performance Indicators as a measure of how well the company do
regarding long-term goals?
Forecast objectives:
● What are the forecasting objectives of company X?
● How is your forecast process regulated by Geneva?
● What requirements do they have?
● How many containers can you store in each terminal before you have to start paying
storage fees?
● How is the demand distributed between different ports?
Appendix
86
What type of container has the greatest demand in the different ports?
Forecasting approach:
● How do you prepare your forecast?
● What kind of forecast approach do you take?
Forecasting variables:
● What variables do you take into account?
● Do you have large oscillations regarding the actual demand?
● Do bookings fluctuate much from period to period?
● Is inventory management part of your forecasting process?
● How long does your clients keep the container?
Time horizon:
● How long time horizon do you take into consideration?
● If it is short term, do you still take into account the long-term horizon? Do you think
this is necessary?
● How long is the lead time from ordering a container from the continent until you
receive it?
● Are you affected by seasonal variations?
● When do you have peaks? Periods during which you have the most and lowest
bookings?
Demand pattern:
● Do you usually check historical data from previous periods (last year/month) when
you do forecasting?
● If you don't use historical data, how do you do it?
● Have you identified trends or seasonal oscillations during the year? When do these
occur?
● If you identify these patterns, do you take them into account when you conduct your
forecasting process?
Forecasting Method:
Appendix
87
● Which forecasting method do you use?
● Do you combine several methods to get better results?
Evaluation of the forecast method
● What criteria do you have when you decide a particular forecasting method?
(Accuracy, costs, expertise of analyst, the way the forecasts will be used)
● Do you measure the accuracy of your forecasting results? If yes, how do you do it?
Integration:
● Once you conducted a result from the process, what happens thereafter?
● Who do you share the conducted information with?
Interview question for the Export department at company X
● Explain the "general" work of the export department.
● How long before sailing / closing do clients place their bookings?
● Do you get many spot bookings? Does these spot bookings include several
containers or only a few?
● How is the distribution between spot bookings and customers who book in advance?
● How long before sailing is information regarding bookings provided to the logistics
department?
● Is it possible to provide information 3 weeks in advance?
● What is the distribution of bookings between different ports?
● What is the demand for different size and type containers in different ports?
● Do some ports have more spot bookings than others? How is the distribution of spot
bookings between different ports?
● How many bookings are cancelled per week / month?
● What are the most common reasons for cancellations?
● Is it possible for the export department to become more proactive in regard to
customers and bookings in order to get better control on spot bookings?
● Is it possible to involve more planning and customer relationship management so
that the process can become more smooth and one can easily predict demand?
Appendix
88
● Is there room for improvement in the forecasting process on the company X? If yes,
can you give suggestions for change?
Interview question for the Import department at company X
● Explain the general work of the import department?
● How is the distribution of import volumes divided between the ports?
● Do you experience season variation?
● In which periods do you experience seasonal highs and lows?
● How long does a customer normally keep a container before it is returned? Does this
varies between different ports?
● Are some customers allowed to keep import units for longer duration of time?
● Explain the cooperation between imports and logistics department.
● Explain the information flow between the two departments, what kind of information
do you share?
● When do you receive information in regard to number of import units? Can this
information be received sooner?
● If the logistics department experience a deficit of containers, will you push your
customers to return units faster?
● Is there room for improvement in forecasting work? If yes, can you provide
suggestions for changes?
Interview question for the Trade commercial export department at
company X
• Explain the general work of trade commercial export department?
• What type of customer contracts do you have?
• How is the distribution between different customer contracts?
• Do you have many tender contracts?
• How do you negotiate tender contracts? Can you explain the process?
• Are there other type contracts that have a major impact on the forecasting of
container volumes?
Appendix
89
• Explain the distribution of bookings between different ports.
• Do you have any estimates of approximately how many number of containers that
are spot bookings per year?
• How do you prepare for these spot bookings?
• When do you provide the logistics department with relevant information?
• Can you provide this information three weeks before closing?
• Is it possible to be more proactive regarding client relationships in order to access
information in advance?
• Is there room for improvement in the forecasting process on the company X? If yes,
can you give suggestions for change?